From f0da47b6f35b71b7c5e502013fd227a65e4f74c7 Mon Sep 17 00:00:00 2001 From: Lookoff123 Date: Tue, 1 Sep 2026 16:32:47 +0500 Subject: [PATCH 1/3] feat(aimlapi): add AI/ML API provider with catalog sync MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit AI/ML API is an aggregator exposing many creators' models behind one OpenAI-compatible key, so it gets a sync provider rather than a hand-written model set. The public catalog needs no key. Only entries served as openai/chat-completions are taken, and of those only the ones whose output is purely text — an id can appear under several endpoint types, and its image or audio surface is a different product. Pricing is quoted as price-per-N-tokens and keyed by origin (provided / generated / cached), which maps onto input / output / cache_read. 163 of the 338 text-output chat models are published here. The rest are skipped by skippedNotice because the catalog does not yet expose a description or an output limit for them; inventing either would be worse than an absent entry. deleteMissing is off since the catalog lists far more than the chat surface, so one response omitting a model is not proof it is gone. Capability flags the catalog does not report (reasoning, tool_call, structured_output, open_weights) keep any value already in the repo instead of being overwritten with a guess. Co-Authored-By: Claude Opus 5 --- package.json | 1 + packages/core/src/sync/index.ts | 4 + packages/core/src/sync/providers/aimlapi.ts | 206 ++++++++++++++++++ providers/aimlapi/logo.svg | 3 + .../Qwen/Qwen2.5-7B-Instruct-Turbo.toml | 23 ++ .../Qwen/Qwen3-235B-A22B-Thinking-2507.toml | 23 ++ .../Qwen3-Coder-480B-A35B-Instruct-FP8.toml | 23 ++ .../models/alibaba/glm-5.2-fast-preview.toml | 24 ++ providers/aimlapi/models/alibaba/glm-5.2.toml | 24 ++ .../aimlapi/models/alibaba/qwen-max.toml | 24 ++ .../aimlapi/models/alibaba/qwen-plus.toml | 24 ++ .../aimlapi/models/alibaba/qwen-turbo.toml | 24 ++ .../qwen3-235b-a22b-thinking-2507.toml | 22 ++ .../aimlapi/models/alibaba/qwen3-32b.toml | 22 ++ .../qwen3-coder-480b-a35b-instruct.toml | 22 ++ .../models/alibaba/qwen3-max-preview.toml | 23 ++ .../aimlapi/models/alibaba/qwen3-max.toml | 23 ++ .../alibaba/qwen3-next-80b-a3b-instruct.toml | 22 ++ .../alibaba/qwen3-next-80b-a3b-thinking.toml | 22 ++ .../alibaba/qwen3-omni-30b-a3b-captioner.toml | 22 ++ .../models/alibaba/qwen3-vl-32b-instruct.toml | 22 ++ .../models/alibaba/qwen3-vl-32b-thinking.toml | 22 ++ .../models/alibaba/qwen3-vl-flash.toml | 23 ++ .../aimlapi/models/alibaba/qwen3-vl-plus.toml | 22 ++ .../aimlapi/models/alibaba/qwen3.5-flash.toml | 23 ++ .../models/alibaba/qwen3.5-plus-20260218.toml | 23 ++ .../aimlapi/models/alibaba/qwen3.5-plus.toml | 23 ++ .../aimlapi/models/alibaba/qwen3.6-flash.toml | 23 ++ .../models/alibaba/qwen3.6-max-preview.toml | 23 ++ .../aimlapi/models/alibaba/qwen3.6-plus.toml | 23 ++ .../aimlapi/models/alibaba/qwen3.7-max.toml | 24 ++ .../aimlapi/models/alibaba/qwen3.7-plus.toml | 24 ++ .../models/alibaba/qwen3.8-2.4t-a95b.toml | 23 ++ .../aimlapi/models/alibaba/qwen3.8-27b.toml | 22 ++ .../aimlapi/models/alibaba/qwen3.8-flash.toml | 23 ++ .../aimlapi/models/alibaba/qwen3.8-max.toml | 24 ++ .../models/anthracite-org/magnum-v4-72b.toml | 22 ++ .../models/anthropic/claude-fable-5.toml | 24 ++ .../anthropic/claude-haiku-4-5-20251001.toml | 24 ++ .../models/anthropic/claude-haiku-4.5.toml | 24 ++ .../anthropic/claude-opus-4-1-20250805.toml | 24 ++ .../anthropic/claude-opus-4-5-20251101.toml | 24 ++ .../models/anthropic/claude-opus-4-6.toml | 24 ++ .../models/anthropic/claude-opus-4-8.toml | 24 ++ .../models/anthropic/claude-opus-4.1.toml | 24 ++ .../models/anthropic/claude-opus-4.5.toml | 24 ++ .../models/anthropic/claude-opus-4.8.toml | 24 ++ .../models/anthropic/claude-opus-5.toml | 24 ++ .../anthropic/claude-sonnet-4-5-20250929.toml | 24 ++ .../models/anthropic/claude-sonnet-4-6.toml | 24 ++ .../models/anthropic/claude-sonnet-4.5.toml | 24 ++ .../models/anthropic/claude-sonnet-4.6.toml | 24 ++ .../models/anthropic/claude-sonnet-5.toml | 24 ++ .../models/baidu/ernie-4.5-vl-424b-a47b.toml | 23 ++ providers/aimlapi/models/baidu/ernie-5.0.toml | 23 ++ .../models/bytedance-seed/seed-2-1-turbo.toml | 23 ++ .../models/bytedance-seed/seed-2.0-code.toml | 23 ++ .../aimlapi/models/bytedance/seed-1-8.toml | 23 ++ .../aimlapi/models/cohere/command-a.toml | 23 ++ .../models/deepseek/deepseek-chat-v3.1.toml | 24 ++ .../models/deepseek/deepseek-chat.toml | 24 ++ .../deepseek-non-reasoner-v3.1-terminus.toml | 24 ++ .../deepseek-non-thinking-v3.2-exp.toml | 24 ++ .../aimlapi/models/deepseek/deepseek-r1.toml | 24 ++ .../deepseek-reasoner-v3.1-terminus.toml | 24 ++ .../models/deepseek/deepseek-reasoner.toml | 24 ++ .../deepseek/deepseek-thinking-v3.2-exp.toml | 24 ++ .../deepseek-v4-flash-vision-exp.toml | 24 ++ .../models/deepseek/deepseek-v4-flash.toml | 24 ++ .../models/deepseek/deepseek-v4-pro-0813.toml | 24 ++ .../models/deepseek/deepseek-v4-pro.toml | 24 ++ .../google/gemini-2.5-flash-lite-preview.toml | 24 ++ .../models/google/gemini-2.5-flash-lite.toml | 24 ++ .../aimlapi/models/google/gemini-2.5-pro.toml | 24 ++ .../models/google/gemini-3-6-flash.toml | 24 ++ .../models/google/gemini-3-7-flash.toml | 24 ++ .../models/google/gemini-3-flash-preview.toml | 24 ++ .../models/google/gemini-3.1-pro-preview.toml | 24 ++ .../models/google/gemini-3.6-flash.toml | 24 ++ .../models/google/gemini-3.7-flash.toml | 24 ++ .../aimlapi/models/google/gemma-3-12b-it.toml | 24 ++ .../aimlapi/models/google/gemma-3-27b-it.toml | 24 ++ .../aimlapi/models/google/gemma-3-4b-it.toml | 24 ++ .../google/gemma-4-26b-a4b-it-maas.toml | 24 ++ .../models/google/gemma-4-26b-a4b-it.toml | 24 ++ .../models/gryphe/mythomax-l2-13b.toml | 23 ++ .../aimlapi/models/meituan/longcat-2.0.toml | 24 ++ .../Llama-3.3-70B-Instruct-Turbo.toml | 23 ++ .../meta-llama/llama-3.3-70b-versatile.toml | 23 ++ .../models/minimax/MiniMax-Text-01.toml | 23 ++ providers/aimlapi/models/minimax/m1.toml | 23 ++ providers/aimlapi/models/minimax/m2-1.toml | 24 ++ .../aimlapi/models/minimax/m2-5-20260218.toml | 24 ++ .../minimax/m2-5-highspeed-20260218.toml | 24 ++ providers/aimlapi/models/minimax/m2.toml | 23 ++ .../aimlapi/models/minimax/minimax-m3.toml | 24 ++ .../models/mistralai/mistral-nemo.toml | 24 ++ .../aimlapi/models/moonshot/kimi-k2-5.toml | 24 ++ .../moonshot/kimi-k2-7-code-highspeed.toml | 24 ++ .../models/moonshot/kimi-k2-7-code.toml | 24 ++ .../models/nousresearch/hermes-4-405b.toml | 24 ++ .../nvidia/nemotron-3-nano-30b-a3b.toml | 23 ++ .../nvidia/nemotron-3-super-120b-a12b.toml | 23 ++ .../nvidia/nemotron-3-ultra-550b-a55b.toml | 24 ++ .../models/nvidia/nemotron-3.5-lightning.toml | 23 ++ .../models/openai/gpt-4.1-2025-04-14.toml | 24 ++ providers/aimlapi/models/openai/gpt-4.1.toml | 24 ++ .../models/openai/gpt-4o-2024-05-13.toml | 23 ++ .../models/openai/gpt-4o-2024-08-06.toml | 24 ++ .../models/openai/gpt-4o-2024-11-20.toml | 24 ++ providers/aimlapi/models/openai/gpt-4o.toml | 24 ++ providers/aimlapi/models/openai/gpt-5-1.toml | 24 ++ .../models/openai/gpt-5-2-chat-latest.toml | 24 ++ providers/aimlapi/models/openai/gpt-5-2.toml | 24 ++ .../models/openai/gpt-5-2025-08-07.toml | 24 ++ .../models/openai/gpt-5-mini-2025-08-07.toml | 24 ++ .../aimlapi/models/openai/gpt-5-mini.toml | 24 ++ .../models/openai/gpt-5-nano-2025-08-07.toml | 24 ++ .../aimlapi/models/openai/gpt-5-nano.toml | 24 ++ .../aimlapi/models/openai/gpt-5-pro.toml | 23 ++ .../models/openai/gpt-5.1-2025-11-13.toml | 24 ++ .../models/openai/gpt-5.1-codex-mini.toml | 24 ++ .../aimlapi/models/openai/gpt-5.1-codex.toml | 24 ++ .../models/openai/gpt-5.2-2025-12-11.toml | 24 ++ .../models/openai/gpt-5.2-chat-latest.toml | 24 ++ .../aimlapi/models/openai/gpt-5.2-codex.toml | 24 ++ .../aimlapi/models/openai/gpt-5.2-pro.toml | 23 ++ .../aimlapi/models/openai/gpt-5.3-codex.toml | 24 ++ providers/aimlapi/models/openai/gpt-5.toml | 24 ++ .../aimlapi/models/openai/gpt-oss-120b.toml | 24 ++ .../aimlapi/models/openai/gpt-oss-20b.toml | 24 ++ .../aimlapi/models/openai/o3-2025-04-16.toml | 24 ++ .../models/openai/o4-mini-2025-04-16.toml | 24 ++ .../aimlapi/models/perplexity/sonar-pro.toml | 23 ++ .../aimlapi/models/perplexity/sonar.toml | 23 ++ .../aimlapi/models/sakana/fugu-ultra.toml | 24 ++ .../aimlapi/models/sakana/sakana-namazu.toml | 24 ++ .../models/stepfun/step-3.7-flash.toml | 24 ++ .../aimlapi/models/tencent/hy-mt2-1.8b.toml | 23 ++ .../models/tencent/hy-mt2-30b-a3b.toml | 23 ++ .../aimlapi/models/tencent/hy-mt2-7b.toml | 23 ++ .../aimlapi/models/tencent/hy4-preview.toml | 23 ++ .../thinkingmachines/inkling-small.toml | 23 ++ .../aimlapi/models/upstage/solar-pro4.toml | 24 ++ .../aimlapi/models/x-ai/grok-3-beta.toml | 24 ++ .../aimlapi/models/x-ai/grok-3-mini-beta.toml | 24 ++ .../x-ai/grok-4-1-fast-non-reasoning.toml | 24 ++ .../models/x-ai/grok-4-1-fast-reasoning.toml | 24 ++ .../x-ai/grok-4-20-0309-non-reasoning.toml | 24 ++ .../models/x-ai/grok-4-20-0309-reasoning.toml | 24 ++ providers/aimlapi/models/x-ai/grok-4-3.toml | 24 ++ .../x-ai/grok-4-fast-non-reasoning.toml | 24 ++ .../models/x-ai/grok-4-fast-reasoning.toml | 24 ++ .../aimlapi/models/x-ai/grok-code-fast-1.toml | 24 ++ .../aimlapi/models/xiaomi/mimo-v2.5-pro.toml | 24 ++ .../aimlapi/models/xiaomi/mimo-v2.5.toml | 24 ++ .../aimlapi/models/z-ai/glm-5.3-flash.toml | 24 ++ .../aimlapi/models/zhipu/glm-4.5-air.toml | 24 ++ providers/aimlapi/models/zhipu/glm-4.5.toml | 24 ++ providers/aimlapi/models/zhipu/glm-4.6.toml | 24 ++ providers/aimlapi/models/zhipu/glm-4.7.toml | 24 ++ .../models/zhipu/glm-5-2-fast-preview.toml | 24 ++ providers/aimlapi/models/zhipu/glm-5-2.toml | 24 ++ providers/aimlapi/models/zhipu/glm-5-3.toml | 24 ++ providers/aimlapi/models/zhipu/glm-5.2.toml | 24 ++ providers/aimlapi/models/zhipu/glm-5.3.toml | 24 ++ providers/aimlapi/models/zhipu/glm-5.toml | 24 ++ providers/aimlapi/provider.toml | 5 + 168 files changed, 4070 insertions(+) create mode 100755 packages/core/src/sync/providers/aimlapi.ts create mode 100755 providers/aimlapi/logo.svg create mode 100644 providers/aimlapi/models/Qwen/Qwen2.5-7B-Instruct-Turbo.toml create mode 100644 providers/aimlapi/models/Qwen/Qwen3-235B-A22B-Thinking-2507.toml create mode 100644 providers/aimlapi/models/Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8.toml create mode 100644 providers/aimlapi/models/alibaba/glm-5.2-fast-preview.toml create mode 100644 providers/aimlapi/models/alibaba/glm-5.2.toml create mode 100644 providers/aimlapi/models/alibaba/qwen-max.toml create mode 100644 providers/aimlapi/models/alibaba/qwen-plus.toml create mode 100644 providers/aimlapi/models/alibaba/qwen-turbo.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3-235b-a22b-thinking-2507.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3-32b.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3-coder-480b-a35b-instruct.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3-max-preview.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3-max.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3-next-80b-a3b-instruct.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3-next-80b-a3b-thinking.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3-omni-30b-a3b-captioner.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3-vl-32b-instruct.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3-vl-32b-thinking.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3-vl-flash.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3-vl-plus.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3.5-flash.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3.5-plus-20260218.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3.5-plus.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3.6-flash.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3.6-max-preview.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3.6-plus.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3.7-max.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3.7-plus.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3.8-2.4t-a95b.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3.8-27b.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3.8-flash.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3.8-max.toml create mode 100644 providers/aimlapi/models/anthracite-org/magnum-v4-72b.toml create mode 100644 providers/aimlapi/models/anthropic/claude-fable-5.toml create mode 100644 providers/aimlapi/models/anthropic/claude-haiku-4-5-20251001.toml create mode 100644 providers/aimlapi/models/anthropic/claude-haiku-4.5.toml create mode 100644 providers/aimlapi/models/anthropic/claude-opus-4-1-20250805.toml create mode 100644 providers/aimlapi/models/anthropic/claude-opus-4-5-20251101.toml create mode 100644 providers/aimlapi/models/anthropic/claude-opus-4-6.toml create mode 100644 providers/aimlapi/models/anthropic/claude-opus-4-8.toml create mode 100644 providers/aimlapi/models/anthropic/claude-opus-4.1.toml create mode 100644 providers/aimlapi/models/anthropic/claude-opus-4.5.toml create mode 100644 providers/aimlapi/models/anthropic/claude-opus-4.8.toml create mode 100644 providers/aimlapi/models/anthropic/claude-opus-5.toml create mode 100644 providers/aimlapi/models/anthropic/claude-sonnet-4-5-20250929.toml create mode 100644 providers/aimlapi/models/anthropic/claude-sonnet-4-6.toml create mode 100644 providers/aimlapi/models/anthropic/claude-sonnet-4.5.toml create mode 100644 providers/aimlapi/models/anthropic/claude-sonnet-4.6.toml create mode 100644 providers/aimlapi/models/anthropic/claude-sonnet-5.toml create mode 100644 providers/aimlapi/models/baidu/ernie-4.5-vl-424b-a47b.toml create mode 100644 providers/aimlapi/models/baidu/ernie-5.0.toml create mode 100644 providers/aimlapi/models/bytedance-seed/seed-2-1-turbo.toml create mode 100644 providers/aimlapi/models/bytedance-seed/seed-2.0-code.toml create mode 100644 providers/aimlapi/models/bytedance/seed-1-8.toml create mode 100644 providers/aimlapi/models/cohere/command-a.toml create mode 100644 providers/aimlapi/models/deepseek/deepseek-chat-v3.1.toml create mode 100644 providers/aimlapi/models/deepseek/deepseek-chat.toml create mode 100644 providers/aimlapi/models/deepseek/deepseek-non-reasoner-v3.1-terminus.toml create mode 100644 providers/aimlapi/models/deepseek/deepseek-non-thinking-v3.2-exp.toml create mode 100644 providers/aimlapi/models/deepseek/deepseek-r1.toml create mode 100644 providers/aimlapi/models/deepseek/deepseek-reasoner-v3.1-terminus.toml create mode 100644 providers/aimlapi/models/deepseek/deepseek-reasoner.toml create mode 100644 providers/aimlapi/models/deepseek/deepseek-thinking-v3.2-exp.toml create mode 100644 providers/aimlapi/models/deepseek/deepseek-v4-flash-vision-exp.toml create mode 100644 providers/aimlapi/models/deepseek/deepseek-v4-flash.toml create mode 100644 providers/aimlapi/models/deepseek/deepseek-v4-pro-0813.toml create mode 100644 providers/aimlapi/models/deepseek/deepseek-v4-pro.toml create mode 100644 providers/aimlapi/models/google/gemini-2.5-flash-lite-preview.toml create mode 100644 providers/aimlapi/models/google/gemini-2.5-flash-lite.toml create mode 100644 providers/aimlapi/models/google/gemini-2.5-pro.toml create mode 100644 providers/aimlapi/models/google/gemini-3-6-flash.toml create mode 100644 providers/aimlapi/models/google/gemini-3-7-flash.toml create mode 100644 providers/aimlapi/models/google/gemini-3-flash-preview.toml create mode 100644 providers/aimlapi/models/google/gemini-3.1-pro-preview.toml create mode 100644 providers/aimlapi/models/google/gemini-3.6-flash.toml create mode 100644 providers/aimlapi/models/google/gemini-3.7-flash.toml create mode 100644 providers/aimlapi/models/google/gemma-3-12b-it.toml create mode 100644 providers/aimlapi/models/google/gemma-3-27b-it.toml create mode 100644 providers/aimlapi/models/google/gemma-3-4b-it.toml create mode 100644 providers/aimlapi/models/google/gemma-4-26b-a4b-it-maas.toml create mode 100644 providers/aimlapi/models/google/gemma-4-26b-a4b-it.toml create mode 100644 providers/aimlapi/models/gryphe/mythomax-l2-13b.toml create mode 100644 providers/aimlapi/models/meituan/longcat-2.0.toml create mode 100644 providers/aimlapi/models/meta-llama/Llama-3.3-70B-Instruct-Turbo.toml create mode 100644 providers/aimlapi/models/meta-llama/llama-3.3-70b-versatile.toml create mode 100644 providers/aimlapi/models/minimax/MiniMax-Text-01.toml create mode 100644 providers/aimlapi/models/minimax/m1.toml create mode 100644 providers/aimlapi/models/minimax/m2-1.toml create mode 100644 providers/aimlapi/models/minimax/m2-5-20260218.toml create mode 100644 providers/aimlapi/models/minimax/m2-5-highspeed-20260218.toml create mode 100644 providers/aimlapi/models/minimax/m2.toml create mode 100644 providers/aimlapi/models/minimax/minimax-m3.toml create mode 100644 providers/aimlapi/models/mistralai/mistral-nemo.toml create mode 100644 providers/aimlapi/models/moonshot/kimi-k2-5.toml create mode 100644 providers/aimlapi/models/moonshot/kimi-k2-7-code-highspeed.toml create mode 100644 providers/aimlapi/models/moonshot/kimi-k2-7-code.toml create mode 100644 providers/aimlapi/models/nousresearch/hermes-4-405b.toml create mode 100644 providers/aimlapi/models/nvidia/nemotron-3-nano-30b-a3b.toml create mode 100644 providers/aimlapi/models/nvidia/nemotron-3-super-120b-a12b.toml create mode 100644 providers/aimlapi/models/nvidia/nemotron-3-ultra-550b-a55b.toml create mode 100644 providers/aimlapi/models/nvidia/nemotron-3.5-lightning.toml create mode 100644 providers/aimlapi/models/openai/gpt-4.1-2025-04-14.toml create mode 100644 providers/aimlapi/models/openai/gpt-4.1.toml create mode 100644 providers/aimlapi/models/openai/gpt-4o-2024-05-13.toml create mode 100644 providers/aimlapi/models/openai/gpt-4o-2024-08-06.toml create mode 100644 providers/aimlapi/models/openai/gpt-4o-2024-11-20.toml create mode 100644 providers/aimlapi/models/openai/gpt-4o.toml create mode 100644 providers/aimlapi/models/openai/gpt-5-1.toml create mode 100644 providers/aimlapi/models/openai/gpt-5-2-chat-latest.toml create mode 100644 providers/aimlapi/models/openai/gpt-5-2.toml create mode 100644 providers/aimlapi/models/openai/gpt-5-2025-08-07.toml create mode 100644 providers/aimlapi/models/openai/gpt-5-mini-2025-08-07.toml create mode 100644 providers/aimlapi/models/openai/gpt-5-mini.toml create mode 100644 providers/aimlapi/models/openai/gpt-5-nano-2025-08-07.toml create mode 100644 providers/aimlapi/models/openai/gpt-5-nano.toml create mode 100644 providers/aimlapi/models/openai/gpt-5-pro.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.1-2025-11-13.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.1-codex-mini.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.1-codex.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.2-2025-12-11.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.2-chat-latest.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.2-codex.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.2-pro.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.3-codex.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.toml create mode 100644 providers/aimlapi/models/openai/gpt-oss-120b.toml create mode 100644 providers/aimlapi/models/openai/gpt-oss-20b.toml create mode 100644 providers/aimlapi/models/openai/o3-2025-04-16.toml create mode 100644 providers/aimlapi/models/openai/o4-mini-2025-04-16.toml create mode 100644 providers/aimlapi/models/perplexity/sonar-pro.toml create mode 100644 providers/aimlapi/models/perplexity/sonar.toml create mode 100644 providers/aimlapi/models/sakana/fugu-ultra.toml create mode 100644 providers/aimlapi/models/sakana/sakana-namazu.toml create mode 100644 providers/aimlapi/models/stepfun/step-3.7-flash.toml create mode 100644 providers/aimlapi/models/tencent/hy-mt2-1.8b.toml create mode 100644 providers/aimlapi/models/tencent/hy-mt2-30b-a3b.toml create mode 100644 providers/aimlapi/models/tencent/hy-mt2-7b.toml create mode 100644 providers/aimlapi/models/tencent/hy4-preview.toml create mode 100644 providers/aimlapi/models/thinkingmachines/inkling-small.toml create mode 100644 providers/aimlapi/models/upstage/solar-pro4.toml create mode 100644 providers/aimlapi/models/x-ai/grok-3-beta.toml create mode 100644 providers/aimlapi/models/x-ai/grok-3-mini-beta.toml create mode 100644 providers/aimlapi/models/x-ai/grok-4-1-fast-non-reasoning.toml create mode 100644 providers/aimlapi/models/x-ai/grok-4-1-fast-reasoning.toml create mode 100644 providers/aimlapi/models/x-ai/grok-4-20-0309-non-reasoning.toml create mode 100644 providers/aimlapi/models/x-ai/grok-4-20-0309-reasoning.toml create mode 100644 providers/aimlapi/models/x-ai/grok-4-3.toml create mode 100644 providers/aimlapi/models/x-ai/grok-4-fast-non-reasoning.toml create mode 100644 providers/aimlapi/models/x-ai/grok-4-fast-reasoning.toml create mode 100644 providers/aimlapi/models/x-ai/grok-code-fast-1.toml create mode 100644 providers/aimlapi/models/xiaomi/mimo-v2.5-pro.toml create mode 100644 providers/aimlapi/models/xiaomi/mimo-v2.5.toml create mode 100644 providers/aimlapi/models/z-ai/glm-5.3-flash.toml create mode 100644 providers/aimlapi/models/zhipu/glm-4.5-air.toml create mode 100644 providers/aimlapi/models/zhipu/glm-4.5.toml create mode 100644 providers/aimlapi/models/zhipu/glm-4.6.toml create mode 100644 providers/aimlapi/models/zhipu/glm-4.7.toml create mode 100644 providers/aimlapi/models/zhipu/glm-5-2-fast-preview.toml create mode 100644 providers/aimlapi/models/zhipu/glm-5-2.toml create mode 100644 providers/aimlapi/models/zhipu/glm-5-3.toml create mode 100644 providers/aimlapi/models/zhipu/glm-5.2.toml create mode 100644 providers/aimlapi/models/zhipu/glm-5.3.toml create mode 100644 providers/aimlapi/models/zhipu/glm-5.toml create mode 100755 providers/aimlapi/provider.toml diff --git a/package.json b/package.json index 4a7b90393f7..6ce70c88b55 100644 --- a/package.json +++ b/package.json @@ -18,6 +18,7 @@ "test": "bun test", "validate": "bun ./packages/core/script/validate.ts", "compare:migrations": "bun ./packages/core/script/compare-model-migrations.ts", + "aimlapi:sync": "bun ./packages/core/script/sync-models.ts aimlapi", "anthropic:sync": "bun ./packages/core/script/sync-models.ts anthropic", "baseten:sync": "bun ./packages/core/script/sync-models.ts baseten", "deepinfra:sync": "bun ./packages/core/script/sync-models.ts deepinfra", diff --git a/packages/core/src/sync/index.ts b/packages/core/src/sync/index.ts index f08532afa14..71f1c4a74ec 100644 --- a/packages/core/src/sync/index.ts +++ b/packages/core/src/sync/index.ts @@ -5,6 +5,7 @@ import { z } from "zod"; import { AuthoredModel, AuthoredModelShape, ModelMetadata } from "../schema.js"; import { openMissingModelIssues } from "./missing-issues.js"; +import { aimlapi } from "./providers/aimlapi.js"; import { ambient } from "./providers/ambient.js"; import { anthropic } from "./providers/anthropic.js"; import { baseten } from "./providers/baseten.js"; @@ -128,6 +129,7 @@ export interface SyncResult { } export const providers: { + aimlapi: SyncProvider; ambient: SyncProvider; anthropic: SyncProvider; baseten: SyncProvider; @@ -162,6 +164,7 @@ export const providers: { wandb: SyncProvider; xai: SyncProvider; } = { + aimlapi, ambient, anthropic, baseten, @@ -199,6 +202,7 @@ export const providers: { export const groups = { aggregators: [ + "aimlapi", "crossmodel", "edenai", "empiriolabs", diff --git a/packages/core/src/sync/providers/aimlapi.ts b/packages/core/src/sync/providers/aimlapi.ts new file mode 100755 index 00000000000..ebfb255fa26 --- /dev/null +++ b/packages/core/src/sync/providers/aimlapi.ts @@ -0,0 +1,206 @@ +import { z } from "zod"; + +import { inferKimiFamily, ModelFamilyValues } from "../../family.js"; +import type { SyncProvider } from "../index.js"; + +// The public catalog needs no key, and `include` is what turns on the pricing +// and modality blocks this sync depends on. +const API_ENDPOINT = "https://api.aimlapi.com/v1/models?include=pricing,modalities"; + +// AI/ML API serves one id under several endpoint types — a model can be both a +// chat model and, say, an image model. Only the chat surface belongs here. +const CHAT_COMPLETIONS_TYPE = "openai/chat-completions"; + +const PricingUnit = z.object({ + name: z.string().nullish(), + content: z.string().nullish(), + origin: z.string().nullish(), + price: z.number().nullish(), + per: z.number().nullish(), +}).passthrough(); + +const Info = z.object({ + name: z.string().nullish(), + description: z.string().nullish(), + developer: z.string().nullish(), + releasedAt: z.string().nullish(), + contextLength: z.number().int().nonnegative().nullish(), + outputMax: z.number().int().nonnegative().nullish(), +}).passthrough(); + +export const AimlapiModel = z.object({ + id: z.string().min(1), + type: z.string().nullish(), + info: Info.nullish(), + modalities: z.object({ + input: z.array(z.string()).nullish(), + output: z.array(z.string()).nullish(), + }).passthrough().nullish(), + pricing: z.object({ + units: z.array(PricingUnit).nullish(), + }).passthrough().nullish(), +}).passthrough(); + +export const AimlapiResponse = z.object({ + data: z.array(AimlapiModel).min(1), +}).passthrough(); + +export type AimlapiModel = z.infer; + +type Modality = "text" | "audio" | "image" | "video" | "pdf"; + +const MODALITIES = new Set(["text", "audio", "image", "video", "pdf"]); + +function normalizeModalities(values: readonly string[] | null | undefined): Modality[] { + const seen = new Set(); + for (const value of values ?? []) { + const normalized = value.toLowerCase(); + if (MODALITIES.has(normalized)) seen.add(normalized as Modality); + } + if (seen.size === 0) seen.add("text"); + return [...seen]; +} + +/** + * Prices are quoted as `price` per `per` tokens; models.dev stores dollars per + * million. The unit discriminator is `origin`, not `measure`: provided is + * input, generated is output, cached is a cache read. Only text token charges + * are taken — a model's image or audio units are a different surface. + */ +function perMillion( + units: readonly z.infer[], + origin: string, +): number | undefined { + const unit = units.find( + (candidate) => + candidate.name === "token" && candidate.content === "text" && candidate.origin === origin, + ); + if (!unit || unit.price == null || !unit.per) return undefined; + return (unit.price / unit.per) * 1_000_000; +} + +function positive(value: number | null | undefined): number | undefined { + return value != null && value > 0 ? value : undefined; +} + +/** + * `family` is a closed vocabulary of model families (`gpt`, `claude`, `o`, …), + * not the vendor prefix — matching on the prefix produces values the schema + * rejects. Longest match first so `claude-sonnet` wins over `claude`, and `o` + * only matches when a digit follows it, as in `o3`. + */ +function inferFamily(id: string, name: string): string | undefined { + const kimi = inferKimiFamily(id, name); + if (kimi !== undefined) return kimi; + + const target = `${id} ${name}`.toLowerCase(); + return [...ModelFamilyValues] + .sort((a, b) => b.length - a.length) + .find((family) => { + const value = family.toLowerCase().replace(/[.*+?^${}()|[\]\\]/g, "\\$&"); + if (family === "o") { + return new RegExp(`(^|[^a-z0-9])${value}(?=\\d|$|[^a-z0-9])`).test(target); + } + return new RegExp(`(^|[^a-z0-9])${value}(?=$|[^a-z0-9])`).test(target); + }); +} + +function releaseDate(value: string | null | undefined): string | undefined { + if (!value) return undefined; + const match = /^(\d{4}-\d{2}-\d{2})/.exec(value.trim()); + return match?.[1]; +} + +export const aimlapi = { + id: "aimlapi", + name: "AI/ML API", + modelsDir: "providers/aimlapi/models", + // The catalog turns over quickly and lists far more than the chat surface, so + // a local model missing from one response is not proof that it is gone. + deleteMissing: false, + sourceID(model) { + return model.type === CHAT_COMPLETIONS_TYPE ? model.id : undefined; + }, + skippedNotice(ids) { + if (ids.length === 0) return []; + return [ + `${ids.length} AI/ML API chat models were skipped because the catalog does not yet publish a description or an output limit for them, both of which this schema requires.`, + `Skipped remote IDs: ${ids.map((id) => `\`${id}\``).join(", ")}`, + ]; + }, + missingNotice(paths) { + if (paths.length === 0) return []; + return [ + `${paths.length} local AI/ML API models were absent from the catalog and were retained for manual lifecycle review.`, + `Retained local paths: ${paths.map((item) => `\`${item}\``).join(", ")}`, + ]; + }, + async fetchModels() { + const response = await fetch(API_ENDPOINT); + if (!response.ok) { + throw new Error(`AI/ML API request failed: ${response.status} ${response.statusText}`); + } + return response.json(); + }, + parseModels(raw) { + return AimlapiResponse.parse(raw).data; + }, + translateModel(model, context) { + // Chat only. The same id may also appear under an image or audio type; + // those entries are a different product and are skipped silently. + if (model.type !== CHAT_COMPLETIONS_TYPE) return undefined; + + const existing = context.existing(model.id); + const info = model.info ?? {}; + const units = model.pricing?.units ?? []; + + const output = normalizeModalities(model.modalities?.output); + // A chat model whose output is not purely text is a media model riding the + // chat protocol, and does not belong in a chat catalog. + if (output.length !== 1 || output[0] !== "text") return undefined; + + const input = normalizeModalities(model.modalities?.input); + const contextLimit = positive(info.contextLength) ?? existing?.limit?.context; + const outputLimit = positive(info.outputMax) ?? existing?.limit?.output; + const description = info.description?.trim() || existing?.description; + + // The schema requires a description and an output limit. Where the catalog + // publishes neither and no local value exists, the model is skipped rather + // than filled in: an invented description or a guessed limit would be worse + // than an absent entry, and skippedNotice makes the gap visible. + if (!description || outputLimit === undefined) return undefined; + + return { + id: model.id, + model: { + name: info.name?.trim() || existing?.name || model.id, + description, + family: existing?.family ?? inferFamily(model.id, info.name?.trim() ?? ""), + release_date: releaseDate(info.releasedAt) ?? existing?.release_date, + last_updated: releaseDate(info.releasedAt) ?? existing?.last_updated, + // The catalog does not report these capabilities. A value already in the + // repo was put there by someone who checked; a default here would only + // overwrite that with a guess. + attachment: existing?.attachment ?? input.length > 1, + reasoning: existing?.reasoning ?? false, + tool_call: existing?.tool_call ?? false, + structured_output: existing?.structured_output ?? false, + open_weights: existing?.open_weights ?? false, + cost: { + input: perMillion(units, "provided") ?? existing?.cost?.input, + output: perMillion(units, "generated") ?? existing?.cost?.output, + cache_read: perMillion(units, "cached") ?? existing?.cost?.cache_read, + }, + limit: { + context: contextLimit, + input: contextLimit, + output: outputLimit, + }, + modalities: { + input, + output, + }, + }, + }; + }, +} satisfies SyncProvider; diff --git a/providers/aimlapi/logo.svg b/providers/aimlapi/logo.svg new file mode 100755 index 00000000000..5bd69eb5625 --- /dev/null +++ b/providers/aimlapi/logo.svg @@ -0,0 +1,3 @@ + + + diff --git a/providers/aimlapi/models/Qwen/Qwen2.5-7B-Instruct-Turbo.toml b/providers/aimlapi/models/Qwen/Qwen2.5-7B-Instruct-Turbo.toml new file mode 100644 index 00000000000..158a428d08e --- /dev/null +++ b/providers/aimlapi/models/Qwen/Qwen2.5-7B-Instruct-Turbo.toml @@ -0,0 +1,23 @@ +name = "Qwen2.5 7B Instruct Turbo" +description = "Qwen 2.5 7B Instruct Turbo excels in coding and instruction following." +family = "qwen" +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.39 +output = 0.39 + +[limit] +context = 32_000 +input = 32_000 +output = 31_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/Qwen/Qwen3-235B-A22B-Thinking-2507.toml b/providers/aimlapi/models/Qwen/Qwen3-235B-A22B-Thinking-2507.toml new file mode 100644 index 00000000000..4d8c764a966 --- /dev/null +++ b/providers/aimlapi/models/Qwen/Qwen3-235B-A22B-Thinking-2507.toml @@ -0,0 +1,23 @@ +name = "Qwen3 Thinking 2507" +description = "Alibaba's Qwen 3 thinking 2507 is a multilingual reasoning model with knowledge augmentation, and creative capabilities." +family = "qwen" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.845 +output = 3.9 + +[limit] +context = 32_000 +input = 32_000 +output = 16_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8.toml b/providers/aimlapi/models/Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8.toml new file mode 100644 index 00000000000..dc0491c81f8 --- /dev/null +++ b/providers/aimlapi/models/Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8.toml @@ -0,0 +1,23 @@ +name = "Qwen3 Coder" +description = "Alibaba's Qwen 3 Coder is a powerful coding agent." +family = "qwen" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 2.6 +output = 2.6 + +[limit] +context = 262_000 +input = 262_000 +output = 65_536 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/glm-5.2-fast-preview.toml b/providers/aimlapi/models/alibaba/glm-5.2-fast-preview.toml new file mode 100644 index 00000000000..e6cf4b4c3bd --- /dev/null +++ b/providers/aimlapi/models/alibaba/glm-5.2-fast-preview.toml @@ -0,0 +1,24 @@ +name = "GLM 5.2 Fast Preview" +description = "A latency-optimized GLM-5.2 preview with the same 1M-token context window and 1.5–2× higher output throughput for real-time chat, agents, and coding." +family = "glm" +release_date = "2026-07-10" +last_updated = "2026-07-10" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 4.55 +output = 14.3 +cache_read = 0.91 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 131_072 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/glm-5.2.toml b/providers/aimlapi/models/alibaba/glm-5.2.toml new file mode 100644 index 00000000000..f012ea0d537 --- /dev/null +++ b/providers/aimlapi/models/alibaba/glm-5.2.toml @@ -0,0 +1,24 @@ +name = "GLM 5.2" +description = "GLM-5.2 is Zhipu AI flagship MoE LLM with a 1M-token context window, optimized for agentic coding, tool use, reasoning, and long-context tasks." +family = "glm" +release_date = "2026-06-16" +last_updated = "2026-06-16" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.82 +output = 5.72 +cache_read = 0.364 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 131_072 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen-max.toml b/providers/aimlapi/models/alibaba/qwen-max.toml new file mode 100644 index 00000000000..4554f9eba34 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen-max.toml @@ -0,0 +1,24 @@ +name = "Qwen Max" +description = "Qwen Max: powerful AI, excels in stability, competes globally." +family = "qwen" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 2.08 +output = 8.32 +cache_read = 0.416 + +[limit] +context = 32_000 +input = 32_000 +output = 8_192 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen-plus.toml b/providers/aimlapi/models/alibaba/qwen-plus.toml new file mode 100644 index 00000000000..29e5653ccb5 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen-plus.toml @@ -0,0 +1,24 @@ +name = "Qwen Plus" +description = "Qwen-Plus: Alibaba's advanced multilingual model with enhanced reasoning and instruction-following capabilities." +family = "qwen" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.52 +output = 1.56 +cache_read = 0.104 + +[limit] +context = 131_000 +input = 131_000 +output = 16_384 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen-turbo.toml b/providers/aimlapi/models/alibaba/qwen-turbo.toml new file mode 100644 index 00000000000..c1cd95d2ef1 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen-turbo.toml @@ -0,0 +1,24 @@ +name = "Qwen Turbo" +description = "Qwen Turbo: optimizes AI agent speed, integrates with RAG, large context window." +family = "qwen" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.065 +output = 0.26 +cache_read = 0.013 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 16_384 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-235b-a22b-thinking-2507.toml b/providers/aimlapi/models/alibaba/qwen3-235b-a22b-thinking-2507.toml new file mode 100644 index 00000000000..492d06a2607 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3-235b-a22b-thinking-2507.toml @@ -0,0 +1,22 @@ +name = "Qwen3 Thinking 2507" +description = "Alibaba's Qwen 3 thinking 2507 is a multilingual reasoning model with knowledge augmentation, and creative capabilities." +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.299 +output = 2.99 + +[limit] +context = 32_000 +input = 32_000 +output = 16_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-32b.toml b/providers/aimlapi/models/alibaba/qwen3-32b.toml new file mode 100644 index 00000000000..d75ce76e131 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3-32b.toml @@ -0,0 +1,22 @@ +name = "Qwen3 32B" +description = "Alibaba's Qwen 3 32B is a multilingual model with instruction-following capabilities and enhanced reasoning." +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.208 +output = 0.832 + +[limit] +context = 131_000 +input = 131_000 +output = 16_384 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-coder-480b-a35b-instruct.toml b/providers/aimlapi/models/alibaba/qwen3-coder-480b-a35b-instruct.toml new file mode 100644 index 00000000000..af1e1067450 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3-coder-480b-a35b-instruct.toml @@ -0,0 +1,22 @@ +name = "Qwen3 Coder" +description = "Alibaba's Qwen 3 Coder is a powerful coding agent." +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.95 +output = 9.75 + +[limit] +context = 262_000 +input = 262_000 +output = 65_536 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-max-preview.toml b/providers/aimlapi/models/alibaba/qwen3-max-preview.toml new file mode 100644 index 00000000000..f78f0fcb224 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3-max-preview.toml @@ -0,0 +1,23 @@ +name = "Qwen3 Max Preview" +description = "Alibaba's Qwen 3 Max Preview is a multilingual model with instruction-following capabilities and enhanced reasoning." +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.56 +output = 7.8 +cache_read = 0.312 + +[limit] +context = 252_000 +input = 252_000 +output = 32_768 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-max.toml b/providers/aimlapi/models/alibaba/qwen3-max.toml new file mode 100644 index 00000000000..81be3f5718a --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3-max.toml @@ -0,0 +1,23 @@ +name = "Qwen3 Max Instruct" +description = "Alibaba's Qwen 3 Max Instruct is the most powerful model in the Qwen 3 series." +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.56 +output = 7.8 +cache_read = 0.312 + +[limit] +context = 262_144 +input = 262_144 +output = 65_536 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-next-80b-a3b-instruct.toml b/providers/aimlapi/models/alibaba/qwen3-next-80b-a3b-instruct.toml new file mode 100644 index 00000000000..8081cee7649 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3-next-80b-a3b-instruct.toml @@ -0,0 +1,22 @@ +name = "Qwen3 Next 80B A3B Instruct" +description = "Alibaba's Qwen 3 Next 80B A3B Instruct is a multilingual model with enhanced instruction-following capabilities." +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.195 +output = 1.56 + +[limit] +context = 129_024 +input = 129_024 +output = 16_384 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-next-80b-a3b-thinking.toml b/providers/aimlapi/models/alibaba/qwen3-next-80b-a3b-thinking.toml new file mode 100644 index 00000000000..5640e1c0adb --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3-next-80b-a3b-thinking.toml @@ -0,0 +1,22 @@ +name = "Qwen3 Next 80B A3B Thinking" +description = "Alibaba's Qwen 3 Next 80B A3B Thinking is a multilingual reasoning model with advanced chain-of-thought capabilities." +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.195 +output = 1.56 + +[limit] +context = 126_976 +input = 126_976 +output = 81_920 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-omni-30b-a3b-captioner.toml b/providers/aimlapi/models/alibaba/qwen3-omni-30b-a3b-captioner.toml new file mode 100644 index 00000000000..134b6d09af6 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3-omni-30b-a3b-captioner.toml @@ -0,0 +1,22 @@ +name = "Qwen3 Omni Captioner" +description = "Qwen3-Omni-Captioner automatically generates accurate and comprehensive descriptions for complex audio, including speech, ambient sounds, music, and sound effects. The model can identify speaker emotions, musical elements, and is suitable for audio content analysis, security audits, and audio editing. Supports audio input up to 40 minutes (1 second = 25 tokens)." +release_date = "2025-11-14" +last_updated = "2025-11-14" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 4.953 +output = 3.9779999999999998 + +[limit] +context = 65_536 +input = 65_536 +output = 32_768 + +[modalities] +input = ["audio"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-vl-32b-instruct.toml b/providers/aimlapi/models/alibaba/qwen3-vl-32b-instruct.toml new file mode 100644 index 00000000000..cebfcf225b2 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3-vl-32b-instruct.toml @@ -0,0 +1,22 @@ +name = "Qwen3 VL 32B Instruct" +description = "Alibaba's Qwen 3 VL 32B Instruct is a powerful vision-language model with hybrid instruction-following capabilities." +release_date = "2025-11-11" +last_updated = "2025-11-11" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.91 +output = 3.64 + +[limit] +context = 126_000 +input = 126_000 +output = 32_768 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-vl-32b-thinking.toml b/providers/aimlapi/models/alibaba/qwen3-vl-32b-thinking.toml new file mode 100644 index 00000000000..40b6f7e4b14 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3-vl-32b-thinking.toml @@ -0,0 +1,22 @@ +name = "Qwen3 VL 32B Thinking" +description = "Alibaba's Qwen 3 VL 32B Thinking is a powerful vision-language model with hybrid thinking capabilities." +release_date = "2025-11-11" +last_updated = "2025-11-11" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.91 +output = 10.92 + +[limit] +context = 126_000 +input = 126_000 +output = 32_768 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-vl-flash.toml b/providers/aimlapi/models/alibaba/qwen3-vl-flash.toml new file mode 100644 index 00000000000..bb3450bfa3f --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3-vl-flash.toml @@ -0,0 +1,23 @@ +name = "Qwen3 VL Flash" +description = "Alibaba's Qwen 3 VL Flash is a fast and cost-effective vision-language model with hybrid thinking capabilities. Supports image understanding with up to 16,384 tokens per image and high-resolution mode for detailed analysis." +release_date = "2025-10-19" +last_updated = "2025-10-19" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.065 +output = 0.52 +cache_read = 0.065 + +[limit] +context = 262_144 +input = 262_144 +output = 32_768 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-vl-plus.toml b/providers/aimlapi/models/alibaba/qwen3-vl-plus.toml new file mode 100644 index 00000000000..cb26d7e6e87 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3-vl-plus.toml @@ -0,0 +1,22 @@ +name = "Qwen3 VL Plus" +description = "Alibaba's Qwen 3 VL Plus is a powerful vision-language model with hybrid thinking capabilities. Supports image understanding with up to 16,384 tokens per image and high-resolution mode for detailed analysis." +release_date = "2025-10-19" +last_updated = "2025-10-19" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.26 +output = 2.08 + +[limit] +context = 262_144 +input = 262_144 +output = 32_768 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.5-flash.toml b/providers/aimlapi/models/alibaba/qwen3.5-flash.toml new file mode 100644 index 00000000000..82c9f83d08f --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3.5-flash.toml @@ -0,0 +1,23 @@ +name = "Qwen3.5 Flash" +description = "Qwen Max: powerful AI, excels in stability, competes globally." +family = "qwen3.5" +release_date = "2026-04-21" +last_updated = "2026-04-21" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.13 +output = 0.52 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 65_536 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.5-plus-20260218.toml b/providers/aimlapi/models/alibaba/qwen3.5-plus-20260218.toml new file mode 100644 index 00000000000..523acb8c130 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3.5-plus-20260218.toml @@ -0,0 +1,23 @@ +name = "Qwen3.5 Plus" +description = "Qwen3.5-Plus is a commercial large language model by Alibaba Cloud designed for long-context text generation and enterprise-grade conversational AI. Supports up to 1M tokens per request with production-ready API stability." +family = "qwen3.5" +release_date = "2026-02-20" +last_updated = "2026-02-20" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.52 +output = 3.12 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 65_536 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.5-plus.toml b/providers/aimlapi/models/alibaba/qwen3.5-plus.toml new file mode 100644 index 00000000000..523acb8c130 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3.5-plus.toml @@ -0,0 +1,23 @@ +name = "Qwen3.5 Plus" +description = "Qwen3.5-Plus is a commercial large language model by Alibaba Cloud designed for long-context text generation and enterprise-grade conversational AI. Supports up to 1M tokens per request with production-ready API stability." +family = "qwen3.5" +release_date = "2026-02-20" +last_updated = "2026-02-20" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.52 +output = 3.12 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 65_536 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.6-flash.toml b/providers/aimlapi/models/alibaba/qwen3.6-flash.toml new file mode 100644 index 00000000000..7a18f32d8ce --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3.6-flash.toml @@ -0,0 +1,23 @@ +name = "Qwen3.6 Flash" +description = "Qwen3.6-Flash — a lightweight and cost-efficient model by Alibaba Cloud designed for simple tasks and high-throughput applications. It offers fast response times with lower cost." +family = "qwen3.6" +release_date = "2026-04-21" +last_updated = "2026-04-21" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.325 +output = 1.95 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 65_536 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.6-max-preview.toml b/providers/aimlapi/models/alibaba/qwen3.6-max-preview.toml new file mode 100644 index 00000000000..50cc3406ff5 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3.6-max-preview.toml @@ -0,0 +1,23 @@ +name = "Qwen3.6 Max Preview" +description = "Qwen3.6-Max-Preview — a high-capability reasoning model by Alibaba Cloud designed for complex tasks and long-form problem solving. It provides strong reasoning performance and high accuracy for advanced use cases." +family = "qwen3.6" +release_date = "2026-04-21" +last_updated = "2026-04-21" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.65 +output = 3.9 + +[limit] +context = 262_144 +input = 262_144 +output = 65_536 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.6-plus.toml b/providers/aimlapi/models/alibaba/qwen3.6-plus.toml new file mode 100644 index 00000000000..df64485cf70 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3.6-plus.toml @@ -0,0 +1,23 @@ +name = "Qwen3.6 Plus" +description = "Qwen3.6-Plus — a balanced general-purpose model by Alibaba Cloud optimized for performance, cost, and speed. Suitable for a wide range of chat and generation tasks." +family = "qwen3.6" +release_date = "2026-04-21" +last_updated = "2026-04-21" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.52 +output = 3.12 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 65_536 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.7-max.toml b/providers/aimlapi/models/alibaba/qwen3.7-max.toml new file mode 100644 index 00000000000..e90f8474b8a --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3.7-max.toml @@ -0,0 +1,24 @@ +name = "Qwen3.7 Max" +description = "Qwen3.7-Max is Alibaba’s flagship reasoning and agentic LLM optimized for coding, productivity, long-context understanding and autonomous workflows. Text input and text output only. Supports function calling, explicit context cache and advanced reasoning capabilities." +family = "qwen3.7-max" +release_date = "2026-05-21" +last_updated = "2026-05-21" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 2.6 +output = 7.8 +cache_read = 2.6 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 65_536 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.7-plus.toml b/providers/aimlapi/models/alibaba/qwen3.7-plus.toml new file mode 100644 index 00000000000..64173d1ebc5 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3.7-plus.toml @@ -0,0 +1,24 @@ +name = "Qwen3.7-Plus" +description = "Qwen3.7-Plus is a multimodal agent model that unifies vision and language into a single agent foundation. It delivers strong vision-language capabilities while retaining agentic strength in coding, tool use, and productivity workflows — including GUI understanding, visual Q&A, and full-modality input for complex software engineering tasks." +family = "qwen3.7-plus" +release_date = "2026-06-01" +last_updated = "2026-06-01" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.416 +output = 1.664 +cache_read = 0.0832 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 65_536 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.8-2.4t-a95b.toml b/providers/aimlapi/models/alibaba/qwen3.8-2.4t-a95b.toml new file mode 100644 index 00000000000..eedc3ee7545 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3.8-2.4t-a95b.toml @@ -0,0 +1,23 @@ +name = "Qwen3.8 2.4T A95B" +description = "Qwen3.8 2.4T A95B is an open-weight sparse mixture-of-experts model from Qwen and the open-weight variant of Qwen3.8 Max, with 95 billion active parameters out of 2.4 trillion total. Reasoning is always on." +release_date = "2026-08-12" +last_updated = "2026-08-12" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 3.4385 +output = 8.59625 +cache_read = 0.6877 + +[limit] +context = 262_144 +input = 262_144 +output = 52_429 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.8-27b.toml b/providers/aimlapi/models/alibaba/qwen3.8-27b.toml new file mode 100644 index 00000000000..8f9f3b4ff71 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3.8-27b.toml @@ -0,0 +1,22 @@ +name = "Qwen3.8 27B" +description = "Qwen3.8 27B is an open-weight dense vision-language model from Qwen. It accepts text, image and video inputs and is suited for coding, professional workflows, research, multimodal interaction and long-running agent tasks, with flexible thinking that can be tuned per request." +release_date = "2026-08-14" +last_updated = "2026-08-14" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.61893 +output = 4.40128 + +[limit] +context = 262_144 +input = 262_144 +output = 131_072 + +[modalities] +input = ["image", "text", "video"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.8-flash.toml b/providers/aimlapi/models/alibaba/qwen3.8-flash.toml new file mode 100644 index 00000000000..10fdead5776 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3.8-flash.toml @@ -0,0 +1,23 @@ +name = "Qwen3.8 Flash" +description = "Qwen3.8-Flash is the speed- and cost-optimized tier of the Qwen3.8 family, pairing hybrid thinking with high throughput for coding assistance, agentic workflows and visual understanding. Natively multimodal: accepts text, images and video and returns text, with a 1M-token context window. Supports function calling, built-in tools (including web search), structured outputs and context caching." +release_date = "2026-08-26" +last_updated = "2026-08-26" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.208 +output = 0.611 +cache_read = 0.0208 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 131_072 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.8-max.toml b/providers/aimlapi/models/alibaba/qwen3.8-max.toml new file mode 100644 index 00000000000..a13befb62e7 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3.8-max.toml @@ -0,0 +1,24 @@ +name = "Qwen3.8 Max" +description = "Qwen3.8-Max is Alibaba’s flagship reasoning and agentic LLM, positioned as its highest-capability tier for complex multi-step reasoning, coding and autonomous workflows. Natively multimodal: accepts text and images and returns text, with a 1M-token context window. Supports hybrid thinking mode, function calling, built-in tools (including web search) and structured outputs." +family = "qwen3.8-max" +release_date = "2026-08-03" +last_updated = "2026-08-03" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 2.6 +output = 7.8 +cache_read = 0.325 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 131_072 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/anthracite-org/magnum-v4-72b.toml b/providers/aimlapi/models/anthracite-org/magnum-v4-72b.toml new file mode 100644 index 00000000000..515e4cdc003 --- /dev/null +++ b/providers/aimlapi/models/anthracite-org/magnum-v4-72b.toml @@ -0,0 +1,22 @@ +name = "Magnum v4 72B" +description = "Magnum V4 is a powerful language model optimized for high-quality text generation tasks." +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 4.1262 +output = 6.877 + +[limit] +context = 32_000 +input = 32_000 +output = 16_384 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-fable-5.toml b/providers/aimlapi/models/anthropic/claude-fable-5.toml new file mode 100644 index 00000000000..7dd55cee867 --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-fable-5.toml @@ -0,0 +1,24 @@ +name = "Claude Fable 5" +description = "Claude Fable 5 is Anthropic's model designed for complex reasoning and agentic coding workflows. Features a 1M token context window, adaptive thinking, and strong performance across multilingual tasks, long-context handling, and vision." +family = "claude-fable" +release_date = "2026-06-09" +last_updated = "2026-06-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 13 +output = 65 +cache_read = 1.3 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 128_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-haiku-4-5-20251001.toml b/providers/aimlapi/models/anthropic/claude-haiku-4-5-20251001.toml new file mode 100644 index 00000000000..ff1d913416f --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-haiku-4-5-20251001.toml @@ -0,0 +1,24 @@ +name = "Claude 4.5 Haiku" +description = "Claude 4.5 Haiku is a fast, lightweight model optimized for quick responses and high-throughput applications." +family = "claude-haiku" +release_date = "2025-10-16" +last_updated = "2025-10-16" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.3 +output = 6.5 +cache_read = 0.13 + +[limit] +context = 200_000 +input = 200_000 +output = 64_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-haiku-4.5.toml b/providers/aimlapi/models/anthropic/claude-haiku-4.5.toml new file mode 100644 index 00000000000..c347be507bf --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-haiku-4.5.toml @@ -0,0 +1,24 @@ +name = "Claude 4.5 Haiku" +description = "Claude 4.5 Haiku is a fast, lightweight model optimized for quick responses and high-throughput applications." +family = "claude-haiku" +release_date = "2025-10-16" +last_updated = "2025-10-16" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.3754 +output = 6.877 +cache_read = 0.13754 + +[limit] +context = 200_000 +input = 200_000 +output = 64_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-opus-4-1-20250805.toml b/providers/aimlapi/models/anthropic/claude-opus-4-1-20250805.toml new file mode 100644 index 00000000000..3da960db23b --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-opus-4-1-20250805.toml @@ -0,0 +1,24 @@ +name = "Claude 4.1 Opus" +description = "Claude 4.1 Opus is an enhanced version of Claude 4 Opus with improved reasoning and instruction-following capabilities." +family = "claude-opus" +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 19.5 +output = 97.5 +cache_read = 1.95 + +[limit] +context = 200_000 +input = 200_000 +output = 32_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-opus-4-5-20251101.toml b/providers/aimlapi/models/anthropic/claude-opus-4-5-20251101.toml new file mode 100644 index 00000000000..e89d5594097 --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-opus-4-5-20251101.toml @@ -0,0 +1,24 @@ +name = "Claude 4.5 Opus" +description = "Claude 4.5 Opus is Anthropic's flagship model for the most demanding reasoning and analysis tasks." +family = "claude-opus" +release_date = "2025-11-25" +last_updated = "2025-11-25" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 6.5 +output = 32.5 +cache_read = 0.65 + +[limit] +context = 200_000 +input = 200_000 +output = 64_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-opus-4-6.toml b/providers/aimlapi/models/anthropic/claude-opus-4-6.toml new file mode 100644 index 00000000000..fcdb333c25b --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-opus-4-6.toml @@ -0,0 +1,24 @@ +name = "Claude 4.6 Opus" +description = "Claude 4.6 Opus delivers Anthropic's highest intelligence with extended output for complex, long-horizon tasks." +family = "claude-opus" +release_date = "2026-02-06" +last_updated = "2026-02-06" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 6.5 +output = 32.5 +cache_read = 0.65 + +[limit] +context = 200_000 +input = 200_000 +output = 128_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-opus-4-8.toml b/providers/aimlapi/models/anthropic/claude-opus-4-8.toml new file mode 100644 index 00000000000..01bf0f0a88f --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-opus-4-8.toml @@ -0,0 +1,24 @@ +name = "Claude 4.8 Opus" +description = "Claude 4.8 Opus is Anthropic's most capable model designed for complex reasoning and agentic coding workflows. Features a 1M token context window, adaptive thinking, and top-tier performance across multilingual tasks, long-context handling, and vision." +family = "claude-opus" +release_date = "2026-05-28" +last_updated = "2026-05-28" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 6.5 +output = 32.5 +cache_read = 0.65 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 128_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-opus-4.1.toml b/providers/aimlapi/models/anthropic/claude-opus-4.1.toml new file mode 100644 index 00000000000..47786dc4e28 --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-opus-4.1.toml @@ -0,0 +1,24 @@ +name = "Claude 4.1 Opus" +description = "Claude 4.1 Opus is an enhanced version of Claude 4 Opus with improved reasoning and instruction-following capabilities." +family = "claude-opus" +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 20.631 +output = 103.155 +cache_read = 2.0631 + +[limit] +context = 200_000 +input = 200_000 +output = 32_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-opus-4.5.toml b/providers/aimlapi/models/anthropic/claude-opus-4.5.toml new file mode 100644 index 00000000000..8a5cb554636 --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-opus-4.5.toml @@ -0,0 +1,24 @@ +name = "Claude 4.5 Opus" +description = "Claude 4.5 Opus is Anthropic's flagship model for the most demanding reasoning and analysis tasks." +family = "claude-opus" +release_date = "2025-11-25" +last_updated = "2025-11-25" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 6.877 +output = 34.385 +cache_read = 0.6877 + +[limit] +context = 200_000 +input = 200_000 +output = 64_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-opus-4.8.toml b/providers/aimlapi/models/anthropic/claude-opus-4.8.toml new file mode 100644 index 00000000000..f724b9df4f8 --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-opus-4.8.toml @@ -0,0 +1,24 @@ +name = "Claude 4.8 Opus" +description = "Claude 4.8 Opus is Anthropic's most capable model designed for complex reasoning and agentic coding workflows. Features a 1M token context window, adaptive thinking, and top-tier performance across multilingual tasks, long-context handling, and vision." +family = "claude-opus" +release_date = "2026-05-28" +last_updated = "2026-05-28" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 6.877 +output = 34.385 +cache_read = 0.6877 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-opus-5.toml b/providers/aimlapi/models/anthropic/claude-opus-5.toml new file mode 100644 index 00000000000..5f622b5b070 --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-opus-5.toml @@ -0,0 +1,24 @@ +name = "Claude Opus 5" +description = "Claude Opus 5 is ideal for complex agentic coding and enterprise work. Features a 1M token context window, adaptive thinking, and top-tier performance across long-horizon agentic tasks, knowledge work, and vision." +family = "claude-opus" +release_date = "2026-07-24" +last_updated = "2026-07-24" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 6.5 +output = 32.5 +cache_read = 0.65 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 128_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-sonnet-4-5-20250929.toml b/providers/aimlapi/models/anthropic/claude-sonnet-4-5-20250929.toml new file mode 100644 index 00000000000..0d392a2e5a5 --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-sonnet-4-5-20250929.toml @@ -0,0 +1,24 @@ +name = "Claude 4.5 Sonnet" +description = "Claude 4.5 Sonnet offers a balance of intelligence and speed for diverse enterprise and coding tasks." +family = "claude-sonnet" +release_date = "2025-10-01" +last_updated = "2025-10-01" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 3.9 +output = 19.5 +cache_read = 0.39 + +[limit] +context = 200_000 +input = 200_000 +output = 64_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-sonnet-4-6.toml b/providers/aimlapi/models/anthropic/claude-sonnet-4-6.toml new file mode 100644 index 00000000000..4a33fa60413 --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-sonnet-4-6.toml @@ -0,0 +1,24 @@ +name = "Claude 4.6 Sonnet" +description = "Claude 4.6 Sonnet is a highly capable model balancing performance and efficiency for everyday AI tasks." +family = "claude-sonnet" +release_date = "2026-02-18" +last_updated = "2026-02-18" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 3.9 +output = 19.5 +cache_read = 0.39 + +[limit] +context = 200_000 +input = 200_000 +output = 64_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-sonnet-4.5.toml b/providers/aimlapi/models/anthropic/claude-sonnet-4.5.toml new file mode 100644 index 00000000000..4bb067b0250 --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-sonnet-4.5.toml @@ -0,0 +1,24 @@ +name = "Claude 4.5 Sonnet" +description = "Claude 4.5 Sonnet offers a balance of intelligence and speed for diverse enterprise and coding tasks." +family = "claude-sonnet" +release_date = "2025-10-01" +last_updated = "2025-10-01" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 4.1262 +output = 20.631 +cache_read = 0.41262 + +[limit] +context = 200_000 +input = 200_000 +output = 64_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-sonnet-4.6.toml b/providers/aimlapi/models/anthropic/claude-sonnet-4.6.toml new file mode 100644 index 00000000000..17417c590c9 --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-sonnet-4.6.toml @@ -0,0 +1,24 @@ +name = "Claude 4.6 Sonnet" +description = "Claude 4.6 Sonnet is a highly capable model balancing performance and efficiency for everyday AI tasks." +family = "claude-sonnet" +release_date = "2026-02-18" +last_updated = "2026-02-18" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 4.1262 +output = 20.631 +cache_read = 0.41262 + +[limit] +context = 200_000 +input = 200_000 +output = 64_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-sonnet-5.toml b/providers/aimlapi/models/anthropic/claude-sonnet-5.toml new file mode 100644 index 00000000000..07b90a1f06b --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-sonnet-5.toml @@ -0,0 +1,24 @@ +name = "Claude Sonnet 5" +description = "Claude Sonnet 5 is a high-performance model for coding and agentic workflows, delivering near-Opus quality at Sonnet cost. Features a 1M token context window, adaptive thinking, and high-resolution vision." +family = "claude-sonnet" +release_date = "2026-06-30" +last_updated = "2026-06-30" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 2.6 +output = 13 +cache_read = 0.26 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 128_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/baidu/ernie-4.5-vl-424b-a47b.toml b/providers/aimlapi/models/baidu/ernie-4.5-vl-424b-a47b.toml new file mode 100644 index 00000000000..0fd7d23d03e --- /dev/null +++ b/providers/aimlapi/models/baidu/ernie-4.5-vl-424b-a47b.toml @@ -0,0 +1,23 @@ +name = "Ernie 4.5 VL 424B A47B" +description = "High-capacity vision-language Ernie 4.5 VL 424B A47B for demanding multimodal workflows." +family = "ernie" +release_date = "2026-01-12" +last_updated = "2026-01-12" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.5577 +output = 1.677 + +[limit] +context = 123_000 +input = 123_000 +output = 123_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/baidu/ernie-5.0.toml b/providers/aimlapi/models/baidu/ernie-5.0.toml new file mode 100644 index 00000000000..0ced30896df --- /dev/null +++ b/providers/aimlapi/models/baidu/ernie-5.0.toml @@ -0,0 +1,23 @@ +name = "ERNIE 5.0" +description = "ERNIE 5.0 is Baidu’s unified foundation model supporting text generation, visual understanding, and deep-thinking reasoning, with a 128K context window." +family = "ernie" +release_date = "2026-06-30" +last_updated = "2026-06-30" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.82 +output = 7.28 + +[limit] +context = 128_000 +input = 128_000 +output = 65_536 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/bytedance-seed/seed-2-1-turbo.toml b/providers/aimlapi/models/bytedance-seed/seed-2-1-turbo.toml new file mode 100644 index 00000000000..4b754c1c105 --- /dev/null +++ b/providers/aimlapi/models/bytedance-seed/seed-2-1-turbo.toml @@ -0,0 +1,23 @@ +name = "Seed 2.1 Turbo" +description = "ByteDance Seed 2.1 Turbo is a multimodal model for coding and long-horizon agent workflows, accepting text, image and video input across a 256K context window." +family = "seed" +release_date = "2026-08-10" +last_updated = "2026-08-10" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.6877 +output = 3.4385 + +[limit] +context = 262_144 +input = 262_144 +output = 262_144 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/bytedance-seed/seed-2.0-code.toml b/providers/aimlapi/models/bytedance-seed/seed-2.0-code.toml new file mode 100644 index 00000000000..8aeb284b365 --- /dev/null +++ b/providers/aimlapi/models/bytedance-seed/seed-2.0-code.toml @@ -0,0 +1,23 @@ +name = "Seed-2.0-Code" +description = "ByteDance Seed-2.0-Code is optimized for agentic coding — frontend development, multilingual programming tasks and coding-agent workflows — with a 256K context window, multimodal (text/image/video) input and selectable reasoning effort." +family = "seed" +release_date = "2026-07-30" +last_updated = "2026-07-30" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.6877 +output = 4.1262 + +[limit] +context = 262_144 +input = 262_144 +output = 131_072 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/bytedance/seed-1-8.toml b/providers/aimlapi/models/bytedance/seed-1-8.toml new file mode 100644 index 00000000000..44b4b7aa6be --- /dev/null +++ b/providers/aimlapi/models/bytedance/seed-1-8.toml @@ -0,0 +1,23 @@ +name = "Seed 1.8" +description = "ByteDance Seed 1.8 is a large language model with a 256K context window optimized for long-context reasoning and generation." +family = "seed" +release_date = "2026-01-12" +last_updated = "2026-01-12" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.325 +output = 2.6 + +[limit] +context = 256_000 +input = 256_000 +output = 224_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/cohere/command-a.toml b/providers/aimlapi/models/cohere/command-a.toml new file mode 100644 index 00000000000..6733854fea4 --- /dev/null +++ b/providers/aimlapi/models/cohere/command-a.toml @@ -0,0 +1,23 @@ +name = "Command A" +description = "Command A is a powerful LLM with advanced capabilities for enterprise applications." +family = "command-a" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 3.4385 +output = 13.754 + +[limit] +context = 256_000 +input = 256_000 +output = 256_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-chat-v3.1.toml b/providers/aimlapi/models/deepseek/deepseek-chat-v3.1.toml new file mode 100644 index 00000000000..c2f784804bc --- /dev/null +++ b/providers/aimlapi/models/deepseek/deepseek-chat-v3.1.toml @@ -0,0 +1,24 @@ +name = "DeepSeek V3.1" +description = "DeepSeek-V3.1 is the latest advanced LLM with improved architecture and performance across various natural language tasks." +family = "deepseek" +release_date = "2025-08-28" +last_updated = "2025-08-28" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.89401 +output = 2.33818 +cache_read = 0.75647 + +[limit] +context = 128_000 +input = 128_000 +output = 8_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-chat.toml b/providers/aimlapi/models/deepseek/deepseek-chat.toml new file mode 100644 index 00000000000..1136d2f3adb --- /dev/null +++ b/providers/aimlapi/models/deepseek/deepseek-chat.toml @@ -0,0 +1,24 @@ +name = "DeepSeek V3" +description = "DeepSeek-V3 is an advanced LLM with efficient architecture and high performance across various natural language tasks." +family = "deepseek" +release_date = "2025-08-28" +last_updated = "2025-08-28" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.364 +output = 0.546 +cache_read = 0.0364 + +[limit] +context = 128_000 +input = 128_000 +output = 124_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-non-reasoner-v3.1-terminus.toml b/providers/aimlapi/models/deepseek/deepseek-non-reasoner-v3.1-terminus.toml new file mode 100644 index 00000000000..ff1f0cc6afc --- /dev/null +++ b/providers/aimlapi/models/deepseek/deepseek-non-reasoner-v3.1-terminus.toml @@ -0,0 +1,24 @@ +name = "DeepSeek V3.1" +description = "DeepSeek-V3.1 is the latest advanced LLM with improved architecture and performance across various natural language tasks." +family = "deepseek" +release_date = "2025-08-28" +last_updated = "2025-08-28" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.371358 +output = 1.3754 +cache_read = 0.371358 + +[limit] +context = 128_000 +input = 128_000 +output = 8_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-non-thinking-v3.2-exp.toml b/providers/aimlapi/models/deepseek/deepseek-non-thinking-v3.2-exp.toml new file mode 100644 index 00000000000..e283befe929 --- /dev/null +++ b/providers/aimlapi/models/deepseek/deepseek-non-thinking-v3.2-exp.toml @@ -0,0 +1,24 @@ +name = "DeepSeek V3.2 Exp" +description = "DeepSeek-V3.2-Exp is an experimental variant of DeepSeek-V3.2 with extended context and non-thinking mode for fast responses." +family = "deepseek" +release_date = "2025-10-03" +last_updated = "2025-10-03" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.371358 +output = 0.563914 +cache_read = 0.371358 + +[limit] +context = 128_000 +input = 128_000 +output = 64_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-r1.toml b/providers/aimlapi/models/deepseek/deepseek-r1.toml new file mode 100644 index 00000000000..0bcdc0d420f --- /dev/null +++ b/providers/aimlapi/models/deepseek/deepseek-r1.toml @@ -0,0 +1,24 @@ +name = "DeepSeek Reasoner V3.1" +description = "DeepSeek Reasoner V3.1 excels in reasoning tasks with advanced chain-of-thought processing and the latest improvements in parameter activation." +family = "deepseek" +release_date = "2025-08-28" +last_updated = "2025-08-28" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.6877 +output = 2.95711 +cache_read = 0.6877 + +[limit] +context = 128_000 +input = 128_000 +output = 64_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-reasoner-v3.1-terminus.toml b/providers/aimlapi/models/deepseek/deepseek-reasoner-v3.1-terminus.toml new file mode 100644 index 00000000000..ff1f0cc6afc --- /dev/null +++ b/providers/aimlapi/models/deepseek/deepseek-reasoner-v3.1-terminus.toml @@ -0,0 +1,24 @@ +name = "DeepSeek V3.1" +description = "DeepSeek-V3.1 is the latest advanced LLM with improved architecture and performance across various natural language tasks." +family = "deepseek" +release_date = "2025-08-28" +last_updated = "2025-08-28" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.371358 +output = 1.3754 +cache_read = 0.371358 + +[limit] +context = 128_000 +input = 128_000 +output = 8_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-reasoner.toml b/providers/aimlapi/models/deepseek/deepseek-reasoner.toml new file mode 100644 index 00000000000..eee838b7579 --- /dev/null +++ b/providers/aimlapi/models/deepseek/deepseek-reasoner.toml @@ -0,0 +1,24 @@ +name = "DeepSeek R1" +description = "DeepSeek R1 excels in reasoning tasks with advanced features like chain-of-thought processing and efficient parameter activation" +family = "deepseek" +release_date = "2025-08-28" +last_updated = "2025-08-28" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.364 +output = 0.546 +cache_read = 0.0364 + +[limit] +context = 128_000 +input = 128_000 +output = 127_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-thinking-v3.2-exp.toml b/providers/aimlapi/models/deepseek/deepseek-thinking-v3.2-exp.toml new file mode 100644 index 00000000000..49d2c75a6c6 --- /dev/null +++ b/providers/aimlapi/models/deepseek/deepseek-thinking-v3.2-exp.toml @@ -0,0 +1,24 @@ +name = "DeepSeek V3.2 Exp" +description = "DeepSeek-V3.2-Exp is an experimental variant of DeepSeek-V3.2 with extended context and non-thinking mode for fast responses." +family = "deepseek-thinking" +release_date = "2025-10-03" +last_updated = "2025-10-03" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.371358 +output = 0.563914 +cache_read = 0.371358 + +[limit] +context = 128_000 +input = 128_000 +output = 64_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-v4-flash-vision-exp.toml b/providers/aimlapi/models/deepseek/deepseek-v4-flash-vision-exp.toml new file mode 100644 index 00000000000..e14ec62d5c9 --- /dev/null +++ b/providers/aimlapi/models/deepseek/deepseek-v4-flash-vision-exp.toml @@ -0,0 +1,24 @@ +name = "DeepSeek V4 Flash Vision Exp" +description = "DeepSeek V4 Flash Vision Exp is an experimental vision-enabled build of DeepSeek V4 Flash. It adds image understanding on top of the base model while matching it on text tasks, with a 1M-token context window and thinking mode enabled by default." +family = "deepseek" +release_date = "2026-08-21" +last_updated = "2026-08-21" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.572 +output = 1.716 +cache_read = 0.0182 + +[limit] +context = 1_048_576 +input = 1_048_576 +output = 384_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-v4-flash.toml b/providers/aimlapi/models/deepseek/deepseek-v4-flash.toml new file mode 100644 index 00000000000..822af6bc3cd --- /dev/null +++ b/providers/aimlapi/models/deepseek/deepseek-v4-flash.toml @@ -0,0 +1,24 @@ +name = "DeepSeek V4 Flash" +description = "DeepSeek V4 Flash and Pro are large language models for chat and reasoning tasks with up to 1M context length. They support both standard and thinking modes, enabling efficient responses or deeper reasoning workflows." +family = "deepseek" +release_date = "2026-04-24" +last_updated = "2026-04-24" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.182 +output = 0.364 +cache_read = 0.00364 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 384_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-v4-pro-0813.toml b/providers/aimlapi/models/deepseek/deepseek-v4-pro-0813.toml new file mode 100644 index 00000000000..710bd150065 --- /dev/null +++ b/providers/aimlapi/models/deepseek/deepseek-v4-pro-0813.toml @@ -0,0 +1,24 @@ +name = "DeepSeek V4 Pro 0813" +description = "DeepSeek V4 Pro 0813 is the GA release of DeepSeek V4 Pro — a large-scale mixture-of-experts model for chat and reasoning tasks with 1M context length, supporting both standard and thinking modes." +family = "deepseek" +release_date = "2026-08-12" +last_updated = "2026-08-12" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 2.393196 +output = 4.786392 +cache_read = 0.199433 + +[limit] +context = 1_048_576 +input = 1_048_576 +output = 384_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-v4-pro.toml b/providers/aimlapi/models/deepseek/deepseek-v4-pro.toml new file mode 100644 index 00000000000..0494f280ab4 --- /dev/null +++ b/providers/aimlapi/models/deepseek/deepseek-v4-pro.toml @@ -0,0 +1,24 @@ +name = "DeepSeek V4 Pro" +description = "DeepSeek V4 Flash and Pro are large language models for chat and reasoning tasks with up to 1M context length. They support both standard and thinking modes, enabling efficient responses or deeper reasoning workflows." +family = "deepseek" +release_date = "2026-04-24" +last_updated = "2026-04-24" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.5655 +output = 1.131 +cache_read = 0.0047125 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 384_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-2.5-flash-lite-preview.toml b/providers/aimlapi/models/google/gemini-2.5-flash-lite-preview.toml new file mode 100644 index 00000000000..015b166c9aa --- /dev/null +++ b/providers/aimlapi/models/google/gemini-2.5-flash-lite-preview.toml @@ -0,0 +1,24 @@ +name = "Gemini 2.5 Flash Lite Preview" +description = "Gemini 2.5 Flash Lite Preview is a lightweight AI model developed by Google, optimized for quick responses and efficient processing, making it ideal for tasks requiring minimal latency and resource consumption" +family = "gemini" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.0975 +output = 0.39 +cache_read = 0.0975 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 1_048_576 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-2.5-flash-lite.toml b/providers/aimlapi/models/google/gemini-2.5-flash-lite.toml new file mode 100644 index 00000000000..015b166c9aa --- /dev/null +++ b/providers/aimlapi/models/google/gemini-2.5-flash-lite.toml @@ -0,0 +1,24 @@ +name = "Gemini 2.5 Flash Lite Preview" +description = "Gemini 2.5 Flash Lite Preview is a lightweight AI model developed by Google, optimized for quick responses and efficient processing, making it ideal for tasks requiring minimal latency and resource consumption" +family = "gemini" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.0975 +output = 0.39 +cache_read = 0.0975 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 1_048_576 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-2.5-pro.toml b/providers/aimlapi/models/google/gemini-2.5-pro.toml new file mode 100644 index 00000000000..7729dd62370 --- /dev/null +++ b/providers/aimlapi/models/google/gemini-2.5-pro.toml @@ -0,0 +1,24 @@ +name = "Gemini 2.5 Pro" +description = "Gemini Pro 2.5 is Google's most advanced reasoning AI model designed for complex tasks, showcasing strong reasoning and coding capabilities. It leads various benchmarks, indicating its enhanced performance." +family = "gemini" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.625 +output = 13 +cache_read = 1.625 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 65_536 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-3-6-flash.toml b/providers/aimlapi/models/google/gemini-3-6-flash.toml new file mode 100644 index 00000000000..8326f7dfdf9 --- /dev/null +++ b/providers/aimlapi/models/google/gemini-3-6-flash.toml @@ -0,0 +1,24 @@ +name = "Gemini 3.6 Flash" +description = "Gemini 3.6 Flash is Google’s most intelligent Flash model, balancing speed with frontier intelligence for strong performance on agentic, coding and multimodal tasks, with superior search and grounding." +family = "gemini" +release_date = "2026-07-21" +last_updated = "2026-07-21" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.95 +output = 9.75 +cache_read = 0.195 + +[limit] +context = 1_048_576 +input = 1_048_576 +output = 65_536 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-3-7-flash.toml b/providers/aimlapi/models/google/gemini-3-7-flash.toml new file mode 100644 index 00000000000..d8dbbdc7314 --- /dev/null +++ b/providers/aimlapi/models/google/gemini-3-7-flash.toml @@ -0,0 +1,24 @@ +name = "Gemini 3.7 Flash" +description = "Gemini 3.7 Flash is the high-efficiency Flash model of the Gemini 3 family, with Pro-level agentic capabilities, stronger code generation and terminal execution, and high token efficiency for multi-step multimodal work." +family = "gemini" +release_date = "2026-08-13" +last_updated = "2026-08-13" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.975 +output = 4.875 +cache_read = 0.0975 + +[limit] +context = 1_048_576 +input = 1_048_576 +output = 65_536 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-3-flash-preview.toml b/providers/aimlapi/models/google/gemini-3-flash-preview.toml new file mode 100644 index 00000000000..1005f2ee312 --- /dev/null +++ b/providers/aimlapi/models/google/gemini-3-flash-preview.toml @@ -0,0 +1,24 @@ +name = "Gemini 3 Flash Preview" +description = "Gemini 3 Flash Preview is a fast, efficient variant of Gemini 3 designed for low-latency multimodal applications." +family = "gemini" +release_date = "2025-12-17" +last_updated = "2025-12-17" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.65 +output = 3.9 +cache_read = 0.65 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 65_536 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-3.1-pro-preview.toml b/providers/aimlapi/models/google/gemini-3.1-pro-preview.toml new file mode 100644 index 00000000000..f79176aabe2 --- /dev/null +++ b/providers/aimlapi/models/google/gemini-3.1-pro-preview.toml @@ -0,0 +1,24 @@ +name = "Gemini 3.1 Pro Preview" +description = "A frontier reasoning model optimized for software engineering and agentic workflows with 1M token context." +family = "gemini" +release_date = "2026-02-25" +last_updated = "2026-02-25" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 2.6 +output = 15.6 +cache_read = 0.65 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 65_536 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-3.6-flash.toml b/providers/aimlapi/models/google/gemini-3.6-flash.toml new file mode 100644 index 00000000000..8326f7dfdf9 --- /dev/null +++ b/providers/aimlapi/models/google/gemini-3.6-flash.toml @@ -0,0 +1,24 @@ +name = "Gemini 3.6 Flash" +description = "Gemini 3.6 Flash is Google’s most intelligent Flash model, balancing speed with frontier intelligence for strong performance on agentic, coding and multimodal tasks, with superior search and grounding." +family = "gemini" +release_date = "2026-07-21" +last_updated = "2026-07-21" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.95 +output = 9.75 +cache_read = 0.195 + +[limit] +context = 1_048_576 +input = 1_048_576 +output = 65_536 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-3.7-flash.toml b/providers/aimlapi/models/google/gemini-3.7-flash.toml new file mode 100644 index 00000000000..d8dbbdc7314 --- /dev/null +++ b/providers/aimlapi/models/google/gemini-3.7-flash.toml @@ -0,0 +1,24 @@ +name = "Gemini 3.7 Flash" +description = "Gemini 3.7 Flash is the high-efficiency Flash model of the Gemini 3 family, with Pro-level agentic capabilities, stronger code generation and terminal execution, and high token efficiency for multi-step multimodal work." +family = "gemini" +release_date = "2026-08-13" +last_updated = "2026-08-13" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.975 +output = 4.875 +cache_read = 0.0975 + +[limit] +context = 1_048_576 +input = 1_048_576 +output = 65_536 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/google/gemma-3-12b-it.toml b/providers/aimlapi/models/google/gemma-3-12b-it.toml new file mode 100644 index 00000000000..220d523bfaf --- /dev/null +++ b/providers/aimlapi/models/google/gemma-3-12b-it.toml @@ -0,0 +1,24 @@ +name = "Gemma 3 12B" +description = "Gemma 3 12B is a mid-size open language model from Google offering a balance of capability and efficiency." +family = "gemma" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.06877 +output = 0.20631 +cache_read = 0.055016 + +[limit] +context = 131_000 +input = 131_000 +output = 96_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/google/gemma-3-27b-it.toml b/providers/aimlapi/models/google/gemma-3-27b-it.toml new file mode 100644 index 00000000000..14469a91d3e --- /dev/null +++ b/providers/aimlapi/models/google/gemma-3-27b-it.toml @@ -0,0 +1,24 @@ +name = "Gemma 3 27B" +description = "Gemma 3 27B is Google's largest Gemma open model, delivering strong reasoning and text generation capabilities." +family = "gemma" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.20631 +output = 0.632684 +cache_read = 0.103155 + +[limit] +context = 128_000 +input = 128_000 +output = 131_072 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/google/gemma-3-4b-it.toml b/providers/aimlapi/models/google/gemma-3-4b-it.toml new file mode 100644 index 00000000000..fae7ae3c2ff --- /dev/null +++ b/providers/aimlapi/models/google/gemma-3-4b-it.toml @@ -0,0 +1,24 @@ +name = "Gemma 3 4B" +description = "Gemma 3 4B is a lightweight open language model from Google, suitable for on-device deployment and efficient inference." +family = "gemma" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.06877 +output = 0.13754 +cache_read = 0.055016 + +[limit] +context = 131_000 +input = 131_000 +output = 32_768 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/google/gemma-4-26b-a4b-it-maas.toml b/providers/aimlapi/models/google/gemma-4-26b-a4b-it-maas.toml new file mode 100644 index 00000000000..276126dd742 --- /dev/null +++ b/providers/aimlapi/models/google/gemma-4-26b-a4b-it-maas.toml @@ -0,0 +1,24 @@ +name = "Gemma 4 26B A4B IT" +description = "Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts open model from Google." +family = "gemma" +release_date = "2026-06-23" +last_updated = "2026-06-23" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.195 +output = 0.78 +cache_read = 0.195 + +[limit] +context = 262_144 +input = 262_144 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/google/gemma-4-26b-a4b-it.toml b/providers/aimlapi/models/google/gemma-4-26b-a4b-it.toml new file mode 100644 index 00000000000..953a8582058 --- /dev/null +++ b/providers/aimlapi/models/google/gemma-4-26b-a4b-it.toml @@ -0,0 +1,24 @@ +name = "Gemma 4 26B A4B IT" +description = "Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts open model from Google." +family = "gemma" +release_date = "2026-06-23" +last_updated = "2026-06-23" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.20631 +output = 0.82524 +cache_read = 0.20631 + +[limit] +context = 262_144 +input = 262_144 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/gryphe/mythomax-l2-13b.toml b/providers/aimlapi/models/gryphe/mythomax-l2-13b.toml new file mode 100644 index 00000000000..a713feabe4d --- /dev/null +++ b/providers/aimlapi/models/gryphe/mythomax-l2-13b.toml @@ -0,0 +1,23 @@ +name = "MythoMax 13B" +description = "One of the highest performing and most popular fine-tunes of Llama 2 13B, with rich descriptions and roleplay." +release_date = "2025-12-09" +last_updated = "2025-12-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 2.578875 +output = 2.578875 +cache_read = 2.578875 + +[limit] +context = 4_096 +input = 4_096 +output = 4_096 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/meituan/longcat-2.0.toml b/providers/aimlapi/models/meituan/longcat-2.0.toml new file mode 100644 index 00000000000..fae4399e81d --- /dev/null +++ b/providers/aimlapi/models/meituan/longcat-2.0.toml @@ -0,0 +1,24 @@ +name = "LongCat-2.0" +description = "LongCat-2.0 is Meituan's trillion-parameter agentic MoE model with a 1M-token context window and up to 128K output tokens. It supports native tool calling and multi-step reasoning." +family = "longcat" +release_date = "2026-06-30" +last_updated = "2026-06-30" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.975 +output = 3.8349999999999995 +cache_read = 0.0195 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 128_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/meta-llama/Llama-3.3-70B-Instruct-Turbo.toml b/providers/aimlapi/models/meta-llama/Llama-3.3-70B-Instruct-Turbo.toml new file mode 100644 index 00000000000..6c8422c2c86 --- /dev/null +++ b/providers/aimlapi/models/meta-llama/Llama-3.3-70B-Instruct-Turbo.toml @@ -0,0 +1,23 @@ +name = "Llama 3.3 70B Instruct Turbo" +description = "Meta Llama 3.3 70B Instruct Turbo is an advanced language model optimized for instruction-following tasks with high efficiency and performance." +family = "llama" +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.144 +output = 1.144 + +[limit] +context = 128_000 +input = 128_000 +output = 127_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/meta-llama/llama-3.3-70b-versatile.toml b/providers/aimlapi/models/meta-llama/llama-3.3-70b-versatile.toml new file mode 100644 index 00000000000..179f2db61c9 --- /dev/null +++ b/providers/aimlapi/models/meta-llama/llama-3.3-70b-versatile.toml @@ -0,0 +1,23 @@ +name = "Llama 3.3 70B Versatile" +description = "Llama-3.3-70B-Versatile is Meta's advanced multilingual large language model, optimized for a wide range of natural language processing tasks. With 70 billion parameters, it offers high performance across various benchmarks while maintaining efficiency suitable for diverse applications." +family = "llama" +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.767 +output = 1.027 + +[limit] +context = 131_072 +input = 131_072 +output = 32_768 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/minimax/MiniMax-Text-01.toml b/providers/aimlapi/models/minimax/MiniMax-Text-01.toml new file mode 100644 index 00000000000..132810f4fc0 --- /dev/null +++ b/providers/aimlapi/models/minimax/MiniMax-Text-01.toml @@ -0,0 +1,23 @@ +name = "Text 01" +description = "MiniMax-Text-01 excels in long-context processing with advanced features like hybrid attention mechanisms and open-source accessibility." +family = "minimax" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.26 +output = 1.43 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 40_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/minimax/m1.toml b/providers/aimlapi/models/minimax/m1.toml new file mode 100644 index 00000000000..d95f3f396eb --- /dev/null +++ b/providers/aimlapi/models/minimax/m1.toml @@ -0,0 +1,23 @@ +name = "MiniMax M1" +description = "MiniMax-M1 is a powerful reasoning model with a 1M token context window, delivering strong performance on complex tasks." +family = "minimax" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.52 +output = 2.86 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 40_960 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/minimax/m2-1.toml b/providers/aimlapi/models/minimax/m2-1.toml new file mode 100644 index 00000000000..f1a9cff43da --- /dev/null +++ b/providers/aimlapi/models/minimax/m2-1.toml @@ -0,0 +1,24 @@ +name = "MiniMax M2.1" +description = "MiniMax-M2.1 is an updated Minimax model with a 200K context window optimized for conversational and reasoning tasks." +family = "minimax" +release_date = "2025-12-23" +last_updated = "2025-12-23" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.39 +output = 1.56 +cache_read = 0.39 + +[limit] +context = 204_000 +input = 204_000 +output = 204_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/minimax/m2-5-20260218.toml b/providers/aimlapi/models/minimax/m2-5-20260218.toml new file mode 100644 index 00000000000..818cb58d735 --- /dev/null +++ b/providers/aimlapi/models/minimax/m2-5-20260218.toml @@ -0,0 +1,24 @@ +name = "MiniMax M2.5" +description = "MiniMax-M2.5 is a general-purpose large language model by MiniMax designed for text generation and conversational AI use cases." +family = "minimax" +release_date = "2026-02-19" +last_updated = "2026-02-19" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.39 +output = 1.56 +cache_read = 0.39 + +[limit] +context = 204_800 +input = 204_800 +output = 204_800 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/minimax/m2-5-highspeed-20260218.toml b/providers/aimlapi/models/minimax/m2-5-highspeed-20260218.toml new file mode 100644 index 00000000000..6797e39df8a --- /dev/null +++ b/providers/aimlapi/models/minimax/m2-5-highspeed-20260218.toml @@ -0,0 +1,24 @@ +name = "MiniMax M2.5 Highspeed" +description = "MiniMax-M2.5 Highspeed is a low-latency, high-throughput version of MiniMax-M2.5 optimized for real-time applications." +family = "minimax" +release_date = "2026-02-19" +last_updated = "2026-02-19" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.78 +output = 3.12 +cache_read = 0.78 + +[limit] +context = 204_800 +input = 204_800 +output = 204_800 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/minimax/m2.toml b/providers/aimlapi/models/minimax/m2.toml new file mode 100644 index 00000000000..ad94739969e --- /dev/null +++ b/providers/aimlapi/models/minimax/m2.toml @@ -0,0 +1,23 @@ +name = "MiniMax M2" +description = "MiniMax-M2 is an advanced large language model with 1M token context for extended reasoning and generation." +family = "minimax" +release_date = "2025-11-04" +last_updated = "2025-11-04" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.39 +output = 1.56 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 40_960 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/minimax/minimax-m3.toml b/providers/aimlapi/models/minimax/minimax-m3.toml new file mode 100644 index 00000000000..be258352180 --- /dev/null +++ b/providers/aimlapi/models/minimax/minimax-m3.toml @@ -0,0 +1,24 @@ +name = "MiniMax M3" +description = "MiniMax M3 is a large-context reasoning model from MiniMax. Supports up to 1M context tokens and is optimized for long-document understanding, coding workflows, agent tasks, and complex multi-step reasoning." +family = "minimax-m3" +release_date = "2026-06-01" +last_updated = "2026-06-01" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.39 +output = 1.56 +cache_read = 0.078 + +[limit] +context = 524_288 +input = 524_288 +output = 524_288 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/mistralai/mistral-nemo.toml b/providers/aimlapi/models/mistralai/mistral-nemo.toml new file mode 100644 index 00000000000..8a5e3238993 --- /dev/null +++ b/providers/aimlapi/models/mistralai/mistral-nemo.toml @@ -0,0 +1,24 @@ +name = "Mistral NeMo" +description = "Mistral-Nemo is a powerful multilingual language model with advanced capabilities." +family = "mistral-nemo" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.20631 +output = 0.233818 +cache_read = 0.20631 + +[limit] +context = 128_000 +input = 128_000 +output = 131_072 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/moonshot/kimi-k2-5.toml b/providers/aimlapi/models/moonshot/kimi-k2-5.toml new file mode 100644 index 00000000000..06b66633c34 --- /dev/null +++ b/providers/aimlapi/models/moonshot/kimi-k2-5.toml @@ -0,0 +1,24 @@ +name = "Kimi K2.5" +description = "Kimi k2.5 is a powerful agentic model by Moonshot AI with strong coding and tool-use capabilities." +family = "kimi-k2" +release_date = "2026-02-06" +last_updated = "2026-02-06" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.78 +output = 3.9 +cache_read = 0.13 + +[limit] +context = 262_144 +input = 262_144 +output = 131_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/moonshot/kimi-k2-7-code-highspeed.toml b/providers/aimlapi/models/moonshot/kimi-k2-7-code-highspeed.toml new file mode 100644 index 00000000000..4387f1e0736 --- /dev/null +++ b/providers/aimlapi/models/moonshot/kimi-k2-7-code-highspeed.toml @@ -0,0 +1,24 @@ +name = "Kimi K2.7 Code Highspeed" +description = "High-throughput variant of Kimi K2.7 Code from Moonshot — the same agentic coding model with extended reasoning and tool use, served at higher speed." +family = "kimi-k2" +release_date = "2026-06-22" +last_updated = "2026-06-22" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 2.47 +output = 10.4 +cache_read = 0.49399999999999994 + +[limit] +context = 262_144 +input = 262_144 +output = 260_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/moonshot/kimi-k2-7-code.toml b/providers/aimlapi/models/moonshot/kimi-k2-7-code.toml new file mode 100644 index 00000000000..11d913e2058 --- /dev/null +++ b/providers/aimlapi/models/moonshot/kimi-k2-7-code.toml @@ -0,0 +1,24 @@ +name = "Kimi K2.7 Code" +description = "Agentic coding model from Moonshot with extended reasoning and tool use." +family = "kimi-k2" +release_date = "2026-06-12" +last_updated = "2026-06-12" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.235 +output = 5.2 +cache_read = 0.24699999999999997 + +[limit] +context = 262_144 +input = 262_144 +output = 260_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/nousresearch/hermes-4-405b.toml b/providers/aimlapi/models/nousresearch/hermes-4-405b.toml new file mode 100644 index 00000000000..495fa1936df --- /dev/null +++ b/providers/aimlapi/models/nousresearch/hermes-4-405b.toml @@ -0,0 +1,24 @@ +name = "Hermes 4 405B" +description = "Hermes 4 is a large-scale reasoning model built on Meta-Llama-3.1-405B with hybrid reasoning mode. It can choose to deliberate internally with ... traces or respond directly, offering flexibility between speed and depth. Supports structured outputs, JSON mode, function calling, and tool use." +family = "nousresearch" +release_date = "2025-10-02" +last_updated = "2025-10-02" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.3754 +output = 4.1262 +cache_read = 1.3754 + +[limit] +context = 131_072 +input = 131_072 +output = 16_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/nvidia/nemotron-3-nano-30b-a3b.toml b/providers/aimlapi/models/nvidia/nemotron-3-nano-30b-a3b.toml new file mode 100644 index 00000000000..6f709bd307a --- /dev/null +++ b/providers/aimlapi/models/nvidia/nemotron-3-nano-30b-a3b.toml @@ -0,0 +1,23 @@ +name = "Nemotron 3 Nano 30B A3B" +description = "NVIDIA Nemotron 3 Nano 30B A3B is a lightweight MoE language model from the Nemotron 3 family, designed for efficient and scalable text generation and agent workflows with strong compute efficiency." +family = "nemotron" +release_date = "2026-04-30" +last_updated = "2026-04-30" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.06877 +output = 0.27508 + +[limit] +context = 262_144 +input = 262_144 +output = 228_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/nvidia/nemotron-3-super-120b-a12b.toml b/providers/aimlapi/models/nvidia/nemotron-3-super-120b-a12b.toml new file mode 100644 index 00000000000..896ed2fc7b5 --- /dev/null +++ b/providers/aimlapi/models/nvidia/nemotron-3-super-120b-a12b.toml @@ -0,0 +1,23 @@ +name = "Nemotron 3 Super 120B A12B" +description = "NVIDIA Nemotron 3 Super 120B A12B is the high-capacity MoE variant of the Nemotron 3 family, delivering strong reasoning and long-context performance for demanding text generation and agent workflows while preserving MoE compute efficiency." +family = "nemotron" +release_date = "2026-04-30" +last_updated = "2026-04-30" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.123786 +output = 0.61893 + +[limit] +context = 262_144 +input = 262_144 +output = 262_144 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/nvidia/nemotron-3-ultra-550b-a55b.toml b/providers/aimlapi/models/nvidia/nemotron-3-ultra-550b-a55b.toml new file mode 100644 index 00000000000..cb7b95b096a --- /dev/null +++ b/providers/aimlapi/models/nvidia/nemotron-3-ultra-550b-a55b.toml @@ -0,0 +1,24 @@ +name = "Nemotron 3 Ultra 550B A55B" +description = "NVIDIA Nemotron 3 Ultra is a reasoning and orchestration model built on a hybrid Transformer-Mamba Mixture-of-Experts architecture, optimized for complex reasoning, long-context analysis, and agent workflows with up to 1M context length." +family = "nemotron" +release_date = "2026-06-04" +last_updated = "2026-06-04" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.82524 +output = 4.95144 +cache_read = 0.20631 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 32_768 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/nvidia/nemotron-3.5-lightning.toml b/providers/aimlapi/models/nvidia/nemotron-3.5-lightning.toml new file mode 100644 index 00000000000..716b6fc15b0 --- /dev/null +++ b/providers/aimlapi/models/nvidia/nemotron-3.5-lightning.toml @@ -0,0 +1,23 @@ +name = "Nemotron 3.5 Lightning" +description = "NVIDIA Nemotron 3.5 Lightning is an open mixture-of-experts model with 3B active parameters out of 30B total, built for high-throughput agentic workloads and specialized task execution with up to 1M context length." +family = "nemotron" +release_date = "2026-08-11" +last_updated = "2026-08-11" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0 +output = 0 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 65_536 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-4.1-2025-04-14.toml b/providers/aimlapi/models/openai/gpt-4.1-2025-04-14.toml new file mode 100644 index 00000000000..75228bcff3b --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-4.1-2025-04-14.toml @@ -0,0 +1,24 @@ +name = "GPT-4.1" +description = "GPT-4.1 is OpenAI's large-context language model with up to 1M token input and improved instruction following." +family = "gpt" +release_date = "2025-08-07" +last_updated = "2025-08-07" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 2.6 +output = 10.4 +cache_read = 0.65 + +[limit] +context = 1_047_576 +input = 1_047_576 +output = 32_768 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-4.1.toml b/providers/aimlapi/models/openai/gpt-4.1.toml new file mode 100644 index 00000000000..75228bcff3b --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-4.1.toml @@ -0,0 +1,24 @@ +name = "GPT-4.1" +description = "GPT-4.1 is OpenAI's large-context language model with up to 1M token input and improved instruction following." +family = "gpt" +release_date = "2025-08-07" +last_updated = "2025-08-07" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 2.6 +output = 10.4 +cache_read = 0.65 + +[limit] +context = 1_047_576 +input = 1_047_576 +output = 32_768 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-4o-2024-05-13.toml b/providers/aimlapi/models/openai/gpt-4o-2024-05-13.toml new file mode 100644 index 00000000000..faa36078832 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-4o-2024-05-13.toml @@ -0,0 +1,23 @@ +name = "GPT-4o" +description = "Multimodal AI model by OpenAI enhancing human-computer interaction." +family = "gpt" +release_date = "2025-08-07" +last_updated = "2025-08-07" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 6.5 +output = 19.5 + +[limit] +context = 128_000 +input = 128_000 +output = 16_384 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-4o-2024-08-06.toml b/providers/aimlapi/models/openai/gpt-4o-2024-08-06.toml new file mode 100644 index 00000000000..0e8522113e3 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-4o-2024-08-06.toml @@ -0,0 +1,24 @@ +name = "GPT-4o" +description = "Multimodal AI model by OpenAI enhancing human-computer interaction." +family = "gpt" +release_date = "2025-08-07" +last_updated = "2025-08-07" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 3.25 +output = 13 +cache_read = 1.625 + +[limit] +context = 128_000 +input = 128_000 +output = 16_384 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-4o-2024-11-20.toml b/providers/aimlapi/models/openai/gpt-4o-2024-11-20.toml new file mode 100644 index 00000000000..0e8522113e3 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-4o-2024-11-20.toml @@ -0,0 +1,24 @@ +name = "GPT-4o" +description = "Multimodal AI model by OpenAI enhancing human-computer interaction." +family = "gpt" +release_date = "2025-08-07" +last_updated = "2025-08-07" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 3.25 +output = 13 +cache_read = 1.625 + +[limit] +context = 128_000 +input = 128_000 +output = 16_384 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-4o.toml b/providers/aimlapi/models/openai/gpt-4o.toml new file mode 100644 index 00000000000..0e8522113e3 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-4o.toml @@ -0,0 +1,24 @@ +name = "GPT-4o" +description = "Multimodal AI model by OpenAI enhancing human-computer interaction." +family = "gpt" +release_date = "2025-08-07" +last_updated = "2025-08-07" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 3.25 +output = 13 +cache_read = 1.625 + +[limit] +context = 128_000 +input = 128_000 +output = 16_384 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5-1.toml b/providers/aimlapi/models/openai/gpt-5-1.toml new file mode 100644 index 00000000000..16a1f12c4b2 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5-1.toml @@ -0,0 +1,24 @@ +name = "GPT-5.1" +description = "GPT-5.1 is an incremental update to GPT-5 with improved accuracy and instruction adherence." +family = "gpt" +release_date = "2025-11-14" +last_updated = "2025-11-14" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.625 +output = 13 +cache_read = 0.1625 + +[limit] +context = 400_000 +input = 400_000 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5-2-chat-latest.toml b/providers/aimlapi/models/openai/gpt-5-2-chat-latest.toml new file mode 100644 index 00000000000..cc5e3c7f319 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5-2-chat-latest.toml @@ -0,0 +1,24 @@ +name = "GPT-5.2 Chat Latest" +description = "The latest conversational release in the GPT-5.2 series with improved dialogue quality." +family = "gpt" +release_date = "2025-12-11" +last_updated = "2025-12-11" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 2.275 +output = 18.2 +cache_read = 0.2275 + +[limit] +context = 128_000 +input = 128_000 +output = 16_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5-2.toml b/providers/aimlapi/models/openai/gpt-5-2.toml new file mode 100644 index 00000000000..127a4d72777 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5-2.toml @@ -0,0 +1,24 @@ +name = "GPT-5.2" +description = "GPT-5.2 is OpenAI's next-generation language model with extended context and enhanced performance." +family = "gpt" +release_date = "2025-12-11" +last_updated = "2025-12-11" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 2.275 +output = 18.2 +cache_read = 0.2275 + +[limit] +context = 400_000 +input = 400_000 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5-2025-08-07.toml b/providers/aimlapi/models/openai/gpt-5-2025-08-07.toml new file mode 100644 index 00000000000..23a961267c8 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5-2025-08-07.toml @@ -0,0 +1,24 @@ +name = "GPT-5" +description = "GPT-5 is OpenAI's most advanced language model with a 400K token context window and superior reasoning capabilities." +family = "gpt" +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.625 +output = 13 +cache_read = 0.1625 + +[limit] +context = 400_000 +input = 400_000 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5-mini-2025-08-07.toml b/providers/aimlapi/models/openai/gpt-5-mini-2025-08-07.toml new file mode 100644 index 00000000000..f2fe2f9dd7c --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5-mini-2025-08-07.toml @@ -0,0 +1,24 @@ +name = "GPT-5 Mini" +description = "GPT-5 mini delivers GPT-5 capabilities in a smaller, faster package optimized for cost-sensitive applications." +family = "gpt" +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.325 +output = 2.6 +cache_read = 0.0325 + +[limit] +context = 400_000 +input = 400_000 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5-mini.toml b/providers/aimlapi/models/openai/gpt-5-mini.toml new file mode 100644 index 00000000000..7ffe48a8ab9 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5-mini.toml @@ -0,0 +1,24 @@ +name = "GPT-5 Mini" +description = "GPT-5 mini delivers GPT-5 capabilities in a smaller, faster package optimized for cost-sensitive applications." +family = "gpt" +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.378235 +output = 3.02588 +cache_read = 0.34385 + +[limit] +context = 400_000 +input = 400_000 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5-nano-2025-08-07.toml b/providers/aimlapi/models/openai/gpt-5-nano-2025-08-07.toml new file mode 100644 index 00000000000..82f8edb5710 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5-nano-2025-08-07.toml @@ -0,0 +1,24 @@ +name = "GPT-5 Nano" +description = "GPT-5 nano is the most lightweight GPT-5 variant, built for ultra-low latency use cases." +family = "gpt" +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.065 +output = 0.52 +cache_read = 0.0065 + +[limit] +context = 400_000 +input = 400_000 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5-nano.toml b/providers/aimlapi/models/openai/gpt-5-nano.toml new file mode 100644 index 00000000000..6680551d190 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5-nano.toml @@ -0,0 +1,24 @@ +name = "GPT-5 Nano" +description = "GPT-5 nano is the most lightweight GPT-5 variant, built for ultra-low latency use cases." +family = "gpt" +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.075647 +output = 0.605176 +cache_read = 0.06877 + +[limit] +context = 400_000 +input = 400_000 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5-pro.toml b/providers/aimlapi/models/openai/gpt-5-pro.toml new file mode 100644 index 00000000000..d00e8536048 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5-pro.toml @@ -0,0 +1,23 @@ +name = "GPT-5 Pro" +description = "GPT-5 Pro is the most powerful GPT-5 variant, offering extended output and superior performance for complex tasks." +family = "gpt" +release_date = "2025-11-21" +last_updated = "2025-11-21" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 20.631 +output = 165.048 + +[limit] +context = 400_000 +input = 400_000 +output = 272_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5.1-2025-11-13.toml b/providers/aimlapi/models/openai/gpt-5.1-2025-11-13.toml new file mode 100644 index 00000000000..16a1f12c4b2 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.1-2025-11-13.toml @@ -0,0 +1,24 @@ +name = "GPT-5.1" +description = "GPT-5.1 is an incremental update to GPT-5 with improved accuracy and instruction adherence." +family = "gpt" +release_date = "2025-11-14" +last_updated = "2025-11-14" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.625 +output = 13 +cache_read = 0.1625 + +[limit] +context = 400_000 +input = 400_000 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5.1-codex-mini.toml b/providers/aimlapi/models/openai/gpt-5.1-codex-mini.toml new file mode 100644 index 00000000000..07247594f82 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.1-codex-mini.toml @@ -0,0 +1,24 @@ +name = "GPT-5.1 Codex Mini" +description = "GPT-5.1 Codex Mini is a lighter coding model optimized for fast code completions and suggestions." +family = "gpt" +release_date = "2025-11-14" +last_updated = "2025-11-14" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.34385 +output = 2.7508 +cache_read = 0.041262 + +[limit] +context = 400_000 +input = 400_000 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5.1-codex.toml b/providers/aimlapi/models/openai/gpt-5.1-codex.toml new file mode 100644 index 00000000000..a4e5580499c --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.1-codex.toml @@ -0,0 +1,24 @@ +name = "GPT-5.1 Codex" +description = "GPT-5.1 Codex is a coding-specialized model from the GPT-5.1 series for code generation and understanding." +family = "gpt" +release_date = "2025-11-14" +last_updated = "2025-11-14" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 3.4385 +output = 27.508 +cache_read = 0.34385 + +[limit] +context = 400_000 +input = 400_000 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5.2-2025-12-11.toml b/providers/aimlapi/models/openai/gpt-5.2-2025-12-11.toml new file mode 100644 index 00000000000..127a4d72777 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.2-2025-12-11.toml @@ -0,0 +1,24 @@ +name = "GPT-5.2" +description = "GPT-5.2 is OpenAI's next-generation language model with extended context and enhanced performance." +family = "gpt" +release_date = "2025-12-11" +last_updated = "2025-12-11" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 2.275 +output = 18.2 +cache_read = 0.2275 + +[limit] +context = 400_000 +input = 400_000 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5.2-chat-latest.toml b/providers/aimlapi/models/openai/gpt-5.2-chat-latest.toml new file mode 100644 index 00000000000..cc5e3c7f319 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.2-chat-latest.toml @@ -0,0 +1,24 @@ +name = "GPT-5.2 Chat Latest" +description = "The latest conversational release in the GPT-5.2 series with improved dialogue quality." +family = "gpt" +release_date = "2025-12-11" +last_updated = "2025-12-11" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 2.275 +output = 18.2 +cache_read = 0.2275 + +[limit] +context = 128_000 +input = 128_000 +output = 16_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5.2-codex.toml b/providers/aimlapi/models/openai/gpt-5.2-codex.toml new file mode 100644 index 00000000000..57ce8a12896 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.2-codex.toml @@ -0,0 +1,24 @@ +name = "GPT-5.2 Codex" +description = "GPT-5.2 Codex is a coding-specialized model from the GPT-5.2 series for advanced code generation." +family = "gpt" +release_date = "2026-01-21" +last_updated = "2026-01-21" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 4.8139 +output = 38.5112 +cache_read = 0.48139 + +[limit] +context = 400_000 +input = 400_000 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5.2-pro.toml b/providers/aimlapi/models/openai/gpt-5.2-pro.toml new file mode 100644 index 00000000000..dcdfb431f96 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.2-pro.toml @@ -0,0 +1,23 @@ +name = "GPT-5.2 Pro" +description = "GPT-5.2 Pro is the flagship GPT-5.2 variant offering maximum output size and reasoning depth." +family = "gpt" +release_date = "2025-12-11" +last_updated = "2025-12-11" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 28.8834 +output = 231.0672 + +[limit] +context = 400_000 +input = 400_000 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5.3-codex.toml b/providers/aimlapi/models/openai/gpt-5.3-codex.toml new file mode 100644 index 00000000000..dfe1a91d921 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.3-codex.toml @@ -0,0 +1,24 @@ +name = "GPT-5.3 Codex" +description = "GPT-5.3 Codex is a coding-specialized model from the GPT-5.3 series for advanced code generation." +family = "gpt" +release_date = "2026-03-02" +last_updated = "2026-03-02" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 4.8139 +output = 38.5112 +cache_read = 0.48139 + +[limit] +context = 400_000 +input = 400_000 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5.toml b/providers/aimlapi/models/openai/gpt-5.toml new file mode 100644 index 00000000000..4244e5c135e --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.toml @@ -0,0 +1,24 @@ +name = "GPT-5" +description = "GPT-5 is OpenAI's most advanced language model with a 400K token context window and superior reasoning capabilities." +family = "gpt" +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.891175 +output = 15.1294 +cache_read = 1.71925 + +[limit] +context = 400_000 +input = 400_000 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-oss-120b.toml b/providers/aimlapi/models/openai/gpt-oss-120b.toml new file mode 100644 index 00000000000..3e92a8894ef --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-oss-120b.toml @@ -0,0 +1,24 @@ +name = "GPT OSS 120B" +description = "GPT OSS 120B is the most powerful Open AI open weight model." +family = "gpt-oss" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.48139 +output = 1.30663 +cache_read = 0.48139 + +[limit] +context = 131_000 +input = 131_000 +output = 131_072 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-oss-20b.toml b/providers/aimlapi/models/openai/gpt-oss-20b.toml new file mode 100644 index 00000000000..0fa6aded63b --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-oss-20b.toml @@ -0,0 +1,24 @@ +name = "GPT OSS 20B" +description = "GPT OSS 20B is a powerful Open AI open weight model." +family = "gpt-oss" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.103155 +output = 0.41262 +cache_read = 0.0515775 + +[limit] +context = 131_000 +input = 131_000 +output = 131_072 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/o3-2025-04-16.toml b/providers/aimlapi/models/openai/o3-2025-04-16.toml new file mode 100644 index 00000000000..366eb3bdea7 --- /dev/null +++ b/providers/aimlapi/models/openai/o3-2025-04-16.toml @@ -0,0 +1,24 @@ +name = "o3" +description = "OpenAI o3 is a frontier reasoning model with exceptional performance on complex scientific, math, and coding tasks." +family = "o" +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 2.6 +output = 10.4 +cache_read = 0.65 + +[limit] +context = 200_000 +input = 200_000 +output = 100_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/openai/o4-mini-2025-04-16.toml b/providers/aimlapi/models/openai/o4-mini-2025-04-16.toml new file mode 100644 index 00000000000..399abd13e1d --- /dev/null +++ b/providers/aimlapi/models/openai/o4-mini-2025-04-16.toml @@ -0,0 +1,24 @@ +name = "o4 Mini" +description = "OpenAI o4-mini is a compact reasoning model that delivers strong performance with lower latency and cost." +family = "o" +release_date = "2025-08-07" +last_updated = "2025-08-07" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.43 +output = 5.72 +cache_read = 0.3575 + +[limit] +context = 200_000 +input = 200_000 +output = 100_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/perplexity/sonar-pro.toml b/providers/aimlapi/models/perplexity/sonar-pro.toml new file mode 100644 index 00000000000..5d88a2f9ea4 --- /dev/null +++ b/providers/aimlapi/models/perplexity/sonar-pro.toml @@ -0,0 +1,23 @@ +name = "Sonar Pro" +description = "Perplexity Sonar Pro is an advanced search-augmented language model with deeper reasoning and broader web coverage." +family = "sonar-pro" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 3.9 +output = 19.5 + +[limit] +context = 200_000 +input = 200_000 +output = 100_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/perplexity/sonar.toml b/providers/aimlapi/models/perplexity/sonar.toml new file mode 100644 index 00000000000..036db6a1d8d --- /dev/null +++ b/providers/aimlapi/models/perplexity/sonar.toml @@ -0,0 +1,23 @@ +name = "Sonar" +description = "Perplexity Sonar is a fast search-augmented language model grounded in real-time web data for up-to-date answers." +family = "sonar" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.3 +output = 1.3 + +[limit] +context = 128_000 +input = 128_000 +output = 100_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/sakana/fugu-ultra.toml b/providers/aimlapi/models/sakana/fugu-ultra.toml new file mode 100644 index 00000000000..37ecd1d9ae2 --- /dev/null +++ b/providers/aimlapi/models/sakana/fugu-ultra.toml @@ -0,0 +1,24 @@ +name = "Fugu Ultra" +description = "Fugu Ultra is the higher-performance model in Sakana AI's Fugu family. Rather than a standard monolithic language model, it is a learned multi-agent orchestration system: a model trained to route tasks across a swappable pool of underlying models and to recursively call instances of itself. It is optimized for complex, multi-step reasoning, coding, and agentic workflows, with orchestration tokens billed as standard input/output tokens." +family = "fugu" +release_date = "2026-06-24" +last_updated = "2026-06-24" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 6.877 +output = 41.262 +cache_read = 0.6877 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/sakana/sakana-namazu.toml b/providers/aimlapi/models/sakana/sakana-namazu.toml new file mode 100644 index 00000000000..1aae96e483d --- /dev/null +++ b/providers/aimlapi/models/sakana/sakana-namazu.toml @@ -0,0 +1,24 @@ +name = "Sakana Namazu" +description = "Sakana Namazu is a Japanese-specialized reasoning model from Sakana AI, based on Kimi K2.6 with additional training for Japanese language and business contexts. It is suited for Japanese instruction following and business workflows, supports a 262K context window, and accepts text, image, and file inputs." +family = "sakana-namazu" +release_date = "2026-08-11" +last_updated = "2026-08-11" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.30663 +output = 5.5016 +cache_read = 0.20631 + +[limit] +context = 262_144 +input = 262_144 +output = 65_536 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/stepfun/step-3.7-flash.toml b/providers/aimlapi/models/stepfun/step-3.7-flash.toml new file mode 100644 index 00000000000..06c8efc9b26 --- /dev/null +++ b/providers/aimlapi/models/stepfun/step-3.7-flash.toml @@ -0,0 +1,24 @@ +name = "Step 3.7 Flash" +description = "Step 3.7 Flash is a multimodal Mixture-of-Experts model from StepFun with text, image, and video understanding. It supports a 256K context window and is optimized for fast inference, reasoning, agent workflows, and long-context productivity tasks." +family = "step" +release_date = "2026-06-01" +last_updated = "2026-06-01" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.27508 +output = 1.58171 +cache_read = 0.055016 + +[limit] +context = 256_000 +input = 256_000 +output = 256_000 + +[modalities] +input = ["image", "text", "video"] +output = ["text"] diff --git a/providers/aimlapi/models/tencent/hy-mt2-1.8b.toml b/providers/aimlapi/models/tencent/hy-mt2-1.8b.toml new file mode 100644 index 00000000000..63fb3e2d26f --- /dev/null +++ b/providers/aimlapi/models/tencent/hy-mt2-1.8b.toml @@ -0,0 +1,23 @@ +name = "Hy-MT2-1.8B" +description = "Hy-MT2-1.8B is a compact 1.8B-parameter translation model from Tencent. It supports 33 language pairs and five Chinese dialect and minority-language pairs, with workflows for structured, delimiter-based, contextual, glossary-based, and style-guided translation." +family = "Hy" +release_date = "2026-08-20" +last_updated = "2026-08-20" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.06051759999999999 +output = 0.2434458 + +[limit] +context = 8_192 +input = 8_192 +output = 4_096 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/tencent/hy-mt2-30b-a3b.toml b/providers/aimlapi/models/tencent/hy-mt2-30b-a3b.toml new file mode 100644 index 00000000000..c12eadca9b3 --- /dev/null +++ b/providers/aimlapi/models/tencent/hy-mt2-30b-a3b.toml @@ -0,0 +1,23 @@ +name = "Hy-MT2-30B-A3B" +description = "Hy-MT2-30B-A3B is Tencent's flagship translation model in the Hy-MT2 family. It supports 33 language pairs plus five Chinese dialect and minority-language pairs, with workflows for structured, delimiter-based, contextual and glossary-based translation." +family = "Hy" +release_date = "2026-08-20" +last_updated = "2026-08-20" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.1017796 +output = 0.405743 + +[limit] +context = 8_192 +input = 8_192 +output = 4_096 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/tencent/hy-mt2-7b.toml b/providers/aimlapi/models/tencent/hy-mt2-7b.toml new file mode 100644 index 00000000000..8b2ab3a201f --- /dev/null +++ b/providers/aimlapi/models/tencent/hy-mt2-7b.toml @@ -0,0 +1,23 @@ +name = "Hy-MT2-7B" +description = "Hy-MT2-7B is a 7B-parameter translation model from Tencent. It supports 33 language pairs and five Chinese dialect and minority-language pairs, with workflows for structured, delimiter-based, contextual, glossary-based, and style-guided translation." +family = "Hy" +release_date = "2026-08-19" +last_updated = "2026-08-19" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.1017796 +output = 0.405743 + +[limit] +context = 8_192 +input = 8_192 +output = 4_096 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/tencent/hy4-preview.toml b/providers/aimlapi/models/tencent/hy4-preview.toml new file mode 100644 index 00000000000..e9ca1a5cb9d --- /dev/null +++ b/providers/aimlapi/models/tencent/hy4-preview.toml @@ -0,0 +1,23 @@ +name = "Hy4 preview" +description = "Hy4 preview is a mixture-of-experts model from Tencent with 49B active parameters out of 770B total, built for coding agents, complex tool-use workflows and productivity tasks. It has a 1M-token context window and configurable reasoning effort." +release_date = "2026-08-28" +last_updated = "2026-08-28" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.1470836 +output = 3.4398754 +cache_read = 0.0577668 + +[limit] +context = 1_048_576 +input = 1_048_576 +output = 64_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/thinkingmachines/inkling-small.toml b/providers/aimlapi/models/thinkingmachines/inkling-small.toml new file mode 100644 index 00000000000..650b4a42afb --- /dev/null +++ b/providers/aimlapi/models/thinkingmachines/inkling-small.toml @@ -0,0 +1,23 @@ +name = "Inkling Small" +description = "Inkling Small is an open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 12B active parameters out of 276B total. It is positioned as the smaller, more efficient member of the Inkling family." +release_date = "2026-07-30" +last_updated = "2026-07-30" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.797732 +output = 1.980576 +cache_read = 0.1595464 + +[limit] +context = 524_288 +input = 524_288 +output = 262_144 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/upstage/solar-pro4.toml b/providers/aimlapi/models/upstage/solar-pro4.toml new file mode 100644 index 00000000000..64980444a79 --- /dev/null +++ b/providers/aimlapi/models/upstage/solar-pro4.toml @@ -0,0 +1,24 @@ +name = "Solar Pro 4" +description = "Solar Pro 4 is a large language model from Upstage. It is suited for agentic workflows, office productivity, document-intensive work, and coding." +family = "solar" +release_date = "2026-08-10" +last_updated = "2026-08-10" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.041262 +output = 0.165048 +cache_read = 0.0082524 + +[limit] +context = 524_288 +input = 524_288 +output = 131_072 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-3-beta.toml b/providers/aimlapi/models/x-ai/grok-3-beta.toml new file mode 100644 index 00000000000..9a3320db9b8 --- /dev/null +++ b/providers/aimlapi/models/x-ai/grok-3-beta.toml @@ -0,0 +1,24 @@ +name = "Grok 3 Beta" +description = "Grok 3 Beta is xAI's advanced language model with strong reasoning and broad knowledge capabilities." +family = "grok" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.625 +output = 3.25 +cache_read = 0.26 + +[limit] +context = 131_000 +input = 131_000 +output = 130_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-3-mini-beta.toml b/providers/aimlapi/models/x-ai/grok-3-mini-beta.toml new file mode 100644 index 00000000000..0d390c78b7f --- /dev/null +++ b/providers/aimlapi/models/x-ai/grok-3-mini-beta.toml @@ -0,0 +1,24 @@ +name = "Grok 3 Beta Mini" +description = "Grok 3 Beta Mini is a compact, efficient variant of Grok 3 optimized for speed and cost-effective inference." +family = "grok" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.625 +output = 3.25 +cache_read = 0.26 + +[limit] +context = 131_000 +input = 131_000 +output = 130_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-4-1-fast-non-reasoning.toml b/providers/aimlapi/models/x-ai/grok-4-1-fast-non-reasoning.toml new file mode 100644 index 00000000000..8c6dd4465e2 --- /dev/null +++ b/providers/aimlapi/models/x-ai/grok-4-1-fast-non-reasoning.toml @@ -0,0 +1,24 @@ +name = "Grok 4.1 Fast Non-Reasoning" +description = "Grok 4.1 Fast without reasoning for faster responses and lower latency in high-performance agentic applications." +family = "grok" +release_date = "2025-11-20" +last_updated = "2025-11-20" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.625 +output = 3.25 +cache_read = 0.26 + +[limit] +context = 2_000_000 +input = 2_000_000 +output = 1_999_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-4-1-fast-reasoning.toml b/providers/aimlapi/models/x-ai/grok-4-1-fast-reasoning.toml new file mode 100644 index 00000000000..c3a03b76a6a --- /dev/null +++ b/providers/aimlapi/models/x-ai/grok-4-1-fast-reasoning.toml @@ -0,0 +1,24 @@ +name = "Grok 4.1 Fast Reasoning" +description = "Grok 4.1 Fast is a frontier multimodal model optimized for high-performance agentic tool calling with advanced reasoning capabilities." +family = "grok" +release_date = "2025-11-20" +last_updated = "2025-11-20" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.625 +output = 3.25 +cache_read = 0.26 + +[limit] +context = 2_000_000 +input = 2_000_000 +output = 1_999_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-4-20-0309-non-reasoning.toml b/providers/aimlapi/models/x-ai/grok-4-20-0309-non-reasoning.toml new file mode 100644 index 00000000000..79ee2f9180c --- /dev/null +++ b/providers/aimlapi/models/x-ai/grok-4-20-0309-non-reasoning.toml @@ -0,0 +1,24 @@ +name = "Grok 4.20 Non-Reasoning" +description = "Grok 4.20 Non-Reasoning — a general-purpose language model optimized for standard text generation and conversational tasks, without extended reasoning capabilities." +family = "grok" +release_date = "2026-04-09" +last_updated = "2026-04-09" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.625 +output = 3.25 +cache_read = 0.26 + +[limit] +context = 2_000_000 +input = 2_000_000 +output = 1_999_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-4-20-0309-reasoning.toml b/providers/aimlapi/models/x-ai/grok-4-20-0309-reasoning.toml new file mode 100644 index 00000000000..bb2642d914b --- /dev/null +++ b/providers/aimlapi/models/x-ai/grok-4-20-0309-reasoning.toml @@ -0,0 +1,24 @@ +name = "Grok 4.20 Reasoning" +description = "Grok 4.20 Reasoning — a large language model designed for advanced reasoning tasks. It excels at multi-step problem solving, logical analysis, and complex instruction following." +family = "grok" +release_date = "2026-04-09" +last_updated = "2026-04-09" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.625 +output = 3.25 +cache_read = 0.26 + +[limit] +context = 2_000_000 +input = 2_000_000 +output = 1_999_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-4-3.toml b/providers/aimlapi/models/x-ai/grok-4-3.toml new file mode 100644 index 00000000000..115e62d2432 --- /dev/null +++ b/providers/aimlapi/models/x-ai/grok-4-3.toml @@ -0,0 +1,24 @@ +name = "Grok 4.3" +description = "Grok 4.3 is a large language model by xAI for text generation and reasoning tasks. Supports up to 1M context tokens and is optimized for conversational AI and analytical workflows." +family = "grok" +release_date = "2026-05-05" +last_updated = "2026-05-05" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.625 +output = 3.25 +cache_read = 0.26 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 1_000_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-4-fast-non-reasoning.toml b/providers/aimlapi/models/x-ai/grok-4-fast-non-reasoning.toml new file mode 100644 index 00000000000..7364af4b893 --- /dev/null +++ b/providers/aimlapi/models/x-ai/grok-4-fast-non-reasoning.toml @@ -0,0 +1,24 @@ +name = "Grok 4 Fast Non-Reasoning" +description = "Grok 4 Fast without reasoning for faster responses and lower latency." +family = "grok" +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.625 +output = 3.25 +cache_read = 0.26 + +[limit] +context = 2_000_000 +input = 2_000_000 +output = 1_999_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-4-fast-reasoning.toml b/providers/aimlapi/models/x-ai/grok-4-fast-reasoning.toml new file mode 100644 index 00000000000..016fec90f40 --- /dev/null +++ b/providers/aimlapi/models/x-ai/grok-4-fast-reasoning.toml @@ -0,0 +1,24 @@ +name = "Grok 4 Fast Reasoning" +description = "Grok 4 Fast with always-on reasoning for complex problem-solving tasks." +family = "grok" +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.625 +output = 3.25 +cache_read = 0.26 + +[limit] +context = 2_000_000 +input = 2_000_000 +output = 1_999_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-code-fast-1.toml b/providers/aimlapi/models/x-ai/grok-code-fast-1.toml new file mode 100644 index 00000000000..6ad387bf9bd --- /dev/null +++ b/providers/aimlapi/models/x-ai/grok-code-fast-1.toml @@ -0,0 +1,24 @@ +name = "Grok Code Fast 1" +description = "A speedy and economical reasoning model that excels at agentic coding with function calling and structured outputs." +family = "grok" +release_date = "2025-09-30" +last_updated = "2025-09-30" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.3 +output = 2.6 +cache_read = 0.26 + +[limit] +context = 256_000 +input = 256_000 +output = 255_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/xiaomi/mimo-v2.5-pro.toml b/providers/aimlapi/models/xiaomi/mimo-v2.5-pro.toml new file mode 100644 index 00000000000..9586b604742 --- /dev/null +++ b/providers/aimlapi/models/xiaomi/mimo-v2.5-pro.toml @@ -0,0 +1,24 @@ +name = "MiMo V2.5 Pro" +description = "MiMo-V2.5-Pro — a flagship reasoning model by Xiaomi designed for complex agentic workflows, long-horizon tasks, and advanced software engineering scenarios with a 1M token context window." +family = "mimo-v2.5-pro" +release_date = "2026-04-23" +last_updated = "2026-04-23" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.598299 +output = 1.196598 +cache_read = 0.00495144 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 128_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/xiaomi/mimo-v2.5.toml b/providers/aimlapi/models/xiaomi/mimo-v2.5.toml new file mode 100644 index 00000000000..cc34d0620fa --- /dev/null +++ b/providers/aimlapi/models/xiaomi/mimo-v2.5.toml @@ -0,0 +1,24 @@ +name = "MiMo V2.5" +description = "MiMo-V2.5 — a native omnimodal model by Xiaomi for text, image, audio, and video understanding. It delivers strong agentic performance at lower cost, making it suitable for general-purpose AI applications." +family = "mimo-v2.5" +release_date = "2026-04-23" +last_updated = "2026-04-23" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.192556 +output = 0.385112 +cache_read = 0.0038511200000000004 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 128_000 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/z-ai/glm-5.3-flash.toml b/providers/aimlapi/models/z-ai/glm-5.3-flash.toml new file mode 100644 index 00000000000..55777da55b8 --- /dev/null +++ b/providers/aimlapi/models/z-ai/glm-5.3-flash.toml @@ -0,0 +1,24 @@ +name = "GLM 5.3 Flash" +description = "GLM-5.3 Flash is the low-cost, high-throughput variant of Zhipu AI GLM-5.3, with a 1M-token context window and text, image and video understanding." +family = "glm" +release_date = "2026-08-26" +last_updated = "2026-08-26" +attachment = true +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.0975 +output = 0.325 +cache_read = 0.0195 + +[limit] +context = 1_048_576 +input = 1_048_576 +output = 131_072 + +[modalities] +input = ["image", "text"] +output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-4.5-air.toml b/providers/aimlapi/models/zhipu/glm-4.5-air.toml new file mode 100644 index 00000000000..70f286dae39 --- /dev/null +++ b/providers/aimlapi/models/zhipu/glm-4.5-air.toml @@ -0,0 +1,24 @@ +name = "GLM 4.5 Air" +description = "Lightweight version of GLM-4.5 for cost-effective applications" +family = "glm" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.26 +output = 1.43 +cache_read = 0.039 + +[limit] +context = 128_000 +input = 128_000 +output = 98_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-4.5.toml b/providers/aimlapi/models/zhipu/glm-4.5.toml new file mode 100644 index 00000000000..a0e264a7738 --- /dev/null +++ b/providers/aimlapi/models/zhipu/glm-4.5.toml @@ -0,0 +1,24 @@ +name = "GLM 4.5" +description = "Advanced version of GLM-4.5 with enhanced capabilities and web search integration" +family = "glm" +release_date = "2025-09-09" +last_updated = "2025-09-09" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.78 +output = 2.86 +cache_read = 0.143 + +[limit] +context = 128_000 +input = 128_000 +output = 98_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-4.6.toml b/providers/aimlapi/models/zhipu/glm-4.6.toml new file mode 100644 index 00000000000..f2dd20741b1 --- /dev/null +++ b/providers/aimlapi/models/zhipu/glm-4.6.toml @@ -0,0 +1,24 @@ +name = "GLM 4.6" +description = "Advanced version of GLM-4.6 with web search integration and enhanced capabilities" +family = "glm" +release_date = "2025-10-02" +last_updated = "2025-10-02" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.78 +output = 2.86 +cache_read = 0.143 + +[limit] +context = 200_000 +input = 200_000 +output = 128_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-4.7.toml b/providers/aimlapi/models/zhipu/glm-4.7.toml new file mode 100644 index 00000000000..70394acd2a1 --- /dev/null +++ b/providers/aimlapi/models/zhipu/glm-4.7.toml @@ -0,0 +1,24 @@ +name = "GLM 4.7" +description = "GLM-4.7 is Zhipu AI flagship text LLM optimized for agentic coding and stable multi-step reasoning, supporting long-context workflows." +family = "glm" +release_date = "2025-12-23" +last_updated = "2025-12-23" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 0.78 +output = 2.86 +cache_read = 0.143 + +[limit] +context = 200_000 +input = 200_000 +output = 128_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-5-2-fast-preview.toml b/providers/aimlapi/models/zhipu/glm-5-2-fast-preview.toml new file mode 100644 index 00000000000..e6cf4b4c3bd --- /dev/null +++ b/providers/aimlapi/models/zhipu/glm-5-2-fast-preview.toml @@ -0,0 +1,24 @@ +name = "GLM 5.2 Fast Preview" +description = "A latency-optimized GLM-5.2 preview with the same 1M-token context window and 1.5–2× higher output throughput for real-time chat, agents, and coding." +family = "glm" +release_date = "2026-07-10" +last_updated = "2026-07-10" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 4.55 +output = 14.3 +cache_read = 0.91 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 131_072 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-5-2.toml b/providers/aimlapi/models/zhipu/glm-5-2.toml new file mode 100644 index 00000000000..25dfee5cbaf --- /dev/null +++ b/providers/aimlapi/models/zhipu/glm-5-2.toml @@ -0,0 +1,24 @@ +name = "GLM 5.2" +description = "GLM-5.2 is Zhipu AI flagship MoE LLM with a 1M-token context window, optimized for agentic coding, tool use, reasoning, and long-context tasks." +family = "glm" +release_date = "2026-06-16" +last_updated = "2026-06-16" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.82 +output = 5.72 +cache_read = 0.338 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 131_072 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-5-3.toml b/providers/aimlapi/models/zhipu/glm-5-3.toml new file mode 100644 index 00000000000..39ef4b76da2 --- /dev/null +++ b/providers/aimlapi/models/zhipu/glm-5-3.toml @@ -0,0 +1,24 @@ +name = "GLM 5.3" +description = "GLM-5.3 is Zhipu AI flagship MoE LLM with a 1M-token context window, built for advanced text generation and deep reasoning." +family = "glm" +release_date = "2026-08-17" +last_updated = "2026-08-17" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.82 +output = 5.72 +cache_read = 0.338 + +[limit] +context = 1_024_000 +input = 1_024_000 +output = 131_072 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-5.2.toml b/providers/aimlapi/models/zhipu/glm-5.2.toml new file mode 100644 index 00000000000..25dfee5cbaf --- /dev/null +++ b/providers/aimlapi/models/zhipu/glm-5.2.toml @@ -0,0 +1,24 @@ +name = "GLM 5.2" +description = "GLM-5.2 is Zhipu AI flagship MoE LLM with a 1M-token context window, optimized for agentic coding, tool use, reasoning, and long-context tasks." +family = "glm" +release_date = "2026-06-16" +last_updated = "2026-06-16" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.82 +output = 5.72 +cache_read = 0.338 + +[limit] +context = 1_000_000 +input = 1_000_000 +output = 131_072 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-5.3.toml b/providers/aimlapi/models/zhipu/glm-5.3.toml new file mode 100644 index 00000000000..39ef4b76da2 --- /dev/null +++ b/providers/aimlapi/models/zhipu/glm-5.3.toml @@ -0,0 +1,24 @@ +name = "GLM 5.3" +description = "GLM-5.3 is Zhipu AI flagship MoE LLM with a 1M-token context window, built for advanced text generation and deep reasoning." +family = "glm" +release_date = "2026-08-17" +last_updated = "2026-08-17" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.82 +output = 5.72 +cache_read = 0.338 + +[limit] +context = 1_024_000 +input = 1_024_000 +output = 131_072 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-5.toml b/providers/aimlapi/models/zhipu/glm-5.toml new file mode 100644 index 00000000000..2e6c89f2318 --- /dev/null +++ b/providers/aimlapi/models/zhipu/glm-5.toml @@ -0,0 +1,24 @@ +name = "GLM 5" +description = "GLM-5 is Zhipu AI next-generation 745B MoE LLM for high-quality text generation and reasoning, optimized for chat, long-form content, and instruction-following tasks." +family = "glm" +release_date = "2026-02-12" +last_updated = "2026-02-12" +attachment = false +reasoning = false +tool_call = false +structured_output = false +open_weights = false + +[cost] +input = 1.3 +output = 4.16 +cache_read = 0.26 + +[limit] +context = 200_000 +input = 200_000 +output = 128_000 + +[modalities] +input = ["text"] +output = ["text"] diff --git a/providers/aimlapi/provider.toml b/providers/aimlapi/provider.toml new file mode 100755 index 00000000000..6b7a3f559b9 --- /dev/null +++ b/providers/aimlapi/provider.toml @@ -0,0 +1,5 @@ +name = "AI/ML API" +env = ["AIMLAPI_API_KEY"] +npm = "@ai-sdk/openai-compatible" +api = "https://api.aimlapi.com/v1" +doc = "https://docs.aimlapi.com" From 69cc1c914b324f004fd100a95cf13167f6515ee4 Mon Sep 17 00:00:00 2001 From: Lookoff123 Date: Wed, 2 Sep 2026 15:03:04 +0500 Subject: [PATCH 2/3] fix(aimlapi): reference lab entries with base_model instead of inlining them MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit AI/ML API hosts other people's models and authors none of them, so every entry has to point at the lab file. AGENTS.md makes that a blocker ("Non-lab hosts use base_model"), and the previous full inlines both violated it and drifted from the repository's own data: 65 models disagreed with their lab entry on reasoning, 75 on tool_call, 37 on structured_output, 17 on attachment. openai/gpt-5 shipped here as reasoning = false, tool_call = false while models/openai/gpt-5.toml says both are true. The cause was a fallback of the shape `existing?.reasoning ?? false`. `existing` only ever reads this provider's own directory, which does not exist for a new provider, so the guard never applied and every model was written as false. The entries now carry no capability flags at all, so there is nothing left to contradict. What a file states is now only what is genuinely ours: cost, a limit when it differs from the base, and reasoning_options. The limit no longer sets `input`, because the catalog publishes a context window and an output cap but no input cap, and equating input with the whole context overwrote the lab's correct split (272k in + 128k out inside a 400k window). reasoning_options is required whenever the base reasons, and no lab entry carries it (0 of 245). It is read per model from the API's own published request schema at /docs-json, so the declared control is the one the endpoint actually accepts rather than an assumption. Coverage is 108 of the 338 text-output chat models. 192 are skipped because this repository has no lab entry to point at — authoring complete lab files for them would mean sourcing capability metadata the catalog does not publish — and 38 because their base reasons but the API does not document the control. Both groups are reported through skippedNotice. Verified: bun run validate exits 0 and resolves aimlapi/openai/gpt-5 to reasoning = true, tool_call = true and the lab limit split, with our price. bun test is 268 pass / 4 fail, the same four failures a pristine tree produces. Re-running the sync reports 108 unchanged. Co-Authored-By: Claude Opus 5 --- packages/core/src/sync/providers/aimlapi.ts | 309 +++++++++++------- .../Qwen/Qwen2.5-7B-Instruct-Turbo.toml | 23 -- .../Qwen/Qwen3-235B-A22B-Thinking-2507.toml | 23 -- .../Qwen3-Coder-480B-A35B-Instruct-FP8.toml | 23 -- .../models/alibaba/glm-5.2-fast-preview.toml | 24 -- providers/aimlapi/models/alibaba/glm-5.2.toml | 24 -- .../aimlapi/models/alibaba/qwen-max.toml | 17 +- .../aimlapi/models/alibaba/qwen-plus.toml | 24 -- .../aimlapi/models/alibaba/qwen-turbo.toml | 24 -- .../qwen3-235b-a22b-thinking-2507.toml | 22 -- .../aimlapi/models/alibaba/qwen3-32b.toml | 22 -- .../qwen3-coder-480b-a35b-instruct.toml | 16 +- .../models/alibaba/qwen3-max-preview.toml | 23 -- .../aimlapi/models/alibaba/qwen3-max.toml | 19 +- .../alibaba/qwen3-next-80b-a3b-instruct.toml | 15 +- .../alibaba/qwen3-next-80b-a3b-thinking.toml | 22 -- .../alibaba/qwen3-omni-30b-a3b-captioner.toml | 22 -- .../models/alibaba/qwen3-vl-32b-instruct.toml | 22 -- .../models/alibaba/qwen3-vl-32b-thinking.toml | 22 -- .../models/alibaba/qwen3-vl-flash.toml | 23 -- .../aimlapi/models/alibaba/qwen3-vl-plus.toml | 22 -- .../aimlapi/models/alibaba/qwen3.5-flash.toml | 24 +- .../models/alibaba/qwen3.5-plus-20260218.toml | 23 -- .../aimlapi/models/alibaba/qwen3.5-plus.toml | 23 -- .../aimlapi/models/alibaba/qwen3.6-27b.toml | 13 + .../models/alibaba/qwen3.6-35b-a3b.toml | 13 + .../aimlapi/models/alibaba/qwen3.6-flash.toml | 23 -- .../models/alibaba/qwen3.6-max-preview.toml | 23 -- .../aimlapi/models/alibaba/qwen3.6-plus.toml | 23 -- .../aimlapi/models/alibaba/qwen3.7-max.toml | 24 +- .../aimlapi/models/alibaba/qwen3.7-plus.toml | 24 -- .../models/alibaba/qwen3.8-2.4t-a95b.toml | 20 +- .../aimlapi/models/alibaba/qwen3.8-27b.toml | 20 +- .../aimlapi/models/alibaba/qwen3.8-flash.toml | 23 +- .../aimlapi/models/alibaba/qwen3.8-max.toml | 24 +- .../models/anthracite-org/magnum-v4-72b.toml | 22 -- .../models/anthropic/claude-fable-5.toml | 24 -- .../anthropic/claude-haiku-4-5-20251001.toml | 24 -- .../models/anthropic/claude-haiku-4.5.toml | 24 +- .../anthropic/claude-opus-4-1-20250805.toml | 24 -- .../anthropic/claude-opus-4-5-20251101.toml | 24 -- .../models/anthropic/claude-opus-4-6.toml | 24 -- .../models/anthropic/claude-opus-4-8.toml | 24 -- .../models/anthropic/claude-opus-4.1.toml | 24 +- .../models/anthropic/claude-opus-4.5.toml | 24 +- .../anthropic/claude-opus-4.7-fast.toml | 10 + .../models/anthropic/claude-opus-4.7.toml | 10 + .../anthropic/claude-opus-4.8-fast.toml | 10 + .../models/anthropic/claude-opus-4.8.toml | 24 +- .../models/anthropic/claude-opus-5.toml | 24 -- .../anthropic/claude-sonnet-4-5-20250929.toml | 24 -- .../models/anthropic/claude-sonnet-4-6.toml | 24 -- .../models/anthropic/claude-sonnet-4.5.toml | 24 +- .../models/anthropic/claude-sonnet-4.6.toml | 21 +- .../models/anthropic/claude-sonnet-5.toml | 24 -- .../arcee-ai/trinity-large-thinking.toml | 14 + .../models/baidu/ernie-4.5-vl-424b-a47b.toml | 23 -- providers/aimlapi/models/baidu/ernie-5.0.toml | 23 -- .../models/bytedance-seed/seed-2-1-turbo.toml | 23 -- .../models/bytedance-seed/seed-2.0-code.toml | 24 +- .../models/bytedance-seed/seed-2.0-lite.toml | 13 + .../models/bytedance-seed/seed-2.0-mini.toml | 13 + .../aimlapi/models/bytedance/seed-1-8.toml | 23 -- .../aimlapi/models/cohere/command-a.toml | 23 -- .../models/cohere/command-r-08-2024.toml | 5 + .../models/cohere/command-r-plus-08-2024.toml | 5 + .../models/cohere/command-r7b-12-2024.toml | 5 + .../models/deepseek/deepseek-chat-v3.1.toml | 24 -- .../models/deepseek/deepseek-chat.toml | 16 +- .../deepseek-non-reasoner-v3.1-terminus.toml | 24 -- .../deepseek-non-thinking-v3.2-exp.toml | 24 -- .../aimlapi/models/deepseek/deepseek-r1.toml | 24 -- .../deepseek-reasoner-v3.1-terminus.toml | 24 -- .../models/deepseek/deepseek-reasoner.toml | 24 -- .../deepseek/deepseek-thinking-v3.2-exp.toml | 24 -- .../deepseek-v4-flash-vision-exp.toml | 21 +- .../models/deepseek/deepseek-v4-flash.toml | 24 +- .../models/deepseek/deepseek-v4-pro-0813.toml | 21 +- .../models/deepseek/deepseek-v4-pro.toml | 24 +- .../models/google/gemini-2.5-flash-image.toml | 10 + .../google/gemini-2.5-flash-lite-preview.toml | 24 -- .../models/google/gemini-2.5-flash-lite.toml | 20 +- .../aimlapi/models/google/gemini-2.5-pro.toml | 24 -- .../models/google/gemini-3-6-flash.toml | 24 -- .../models/google/gemini-3-7-flash.toml | 24 -- .../models/google/gemini-3-flash-preview.toml | 24 -- .../google/gemini-3-pro-image-preview.toml | 10 + .../models/google/gemini-3-pro-image.toml | 10 + .../gemini-3.1-flash-image-preview.toml | 13 + .../models/google/gemini-3.1-flash-image.toml | 9 + .../google/gemini-3.1-flash-lite-image.toml | 12 + .../google/gemini-3.1-flash-lite-preview.toml | 13 + .../gemini-3.1-pro-preview-customtools.toml | 13 + .../models/google/gemini-3.1-pro-preview.toml | 24 -- .../models/google/gemini-3.6-flash.toml | 24 +- .../models/google/gemini-3.7-flash.toml | 24 +- .../models/google/gemini-flash-latest.toml | 10 + .../aimlapi/models/google/gemma-3-12b-it.toml | 24 -- .../aimlapi/models/google/gemma-3-27b-it.toml | 24 -- .../aimlapi/models/google/gemma-3-4b-it.toml | 24 -- .../google/gemma-4-26b-a4b-it-maas.toml | 24 -- .../models/google/gemma-4-26b-a4b-it.toml | 21 +- .../aimlapi/models/google/gemma-4-31b-it.toml | 14 + .../models/gryphe/mythomax-l2-13b.toml | 23 -- .../aimlapi/models/meituan/longcat-2.0.toml | 24 -- .../Llama-3.3-70B-Instruct-Turbo.toml | 23 -- .../meta-llama/llama-3.3-70b-versatile.toml | 23 -- .../aimlapi/models/meta/muse-glimmer-30b.toml | 10 + .../aimlapi/models/meta/muse-spark-1.1.toml | 13 + .../aimlapi/models/meta/muse-spark-1.2.toml | 10 + .../models/minimax/MiniMax-Text-01.toml | 23 -- providers/aimlapi/models/minimax/m1.toml | 23 -- providers/aimlapi/models/minimax/m2-1.toml | 24 -- .../aimlapi/models/minimax/m2-5-20260218.toml | 24 -- .../minimax/m2-5-highspeed-20260218.toml | 24 -- providers/aimlapi/models/minimax/m2.toml | 23 -- .../aimlapi/models/minimax/minimax-m3.toml | 24 -- .../models/mistralai/devstral-2512.toml | 6 + .../models/mistralai/mistral-large-2512.toml | 6 + .../models/mistralai/mistral-nemo.toml | 17 +- .../models/mistralai/mistral-small-2603.toml | 13 + .../aimlapi/models/moonshot/kimi-k2-5.toml | 24 -- .../moonshot/kimi-k2-7-code-highspeed.toml | 24 -- .../models/moonshot/kimi-k2-7-code.toml | 24 -- .../models/moonshotai/kimi-k2-thinking.toml | 13 + .../models/nousresearch/hermes-4-405b.toml | 24 -- .../nvidia/nemotron-3-nano-30b-a3b.toml | 23 -- .../nvidia/nemotron-3-super-120b-a12b.toml | 23 -- .../nvidia/nemotron-3-ultra-550b-a55b.toml | 24 -- .../models/nvidia/nemotron-3.5-lightning.toml | 23 -- .../aimlapi/models/openai/gpt-3.5-turbo.toml | 8 + .../aimlapi/models/openai/gpt-4-turbo.toml | 5 + .../models/openai/gpt-4.1-2025-04-14.toml | 24 -- .../aimlapi/models/openai/gpt-4.1-mini.toml | 9 + .../aimlapi/models/openai/gpt-4.1-nano.toml | 9 + providers/aimlapi/models/openai/gpt-4.1.toml | 20 +- providers/aimlapi/models/openai/gpt-4.toml | 9 + .../models/openai/gpt-4o-2024-05-13.toml | 17 +- .../models/openai/gpt-4o-2024-08-06.toml | 20 +- .../models/openai/gpt-4o-2024-11-20.toml | 20 +- .../aimlapi/models/openai/gpt-4o-mini.toml | 6 + providers/aimlapi/models/openai/gpt-4o.toml | 20 +- providers/aimlapi/models/openai/gpt-5-1.toml | 24 -- .../models/openai/gpt-5-2-chat-latest.toml | 24 -- providers/aimlapi/models/openai/gpt-5-2.toml | 24 -- .../models/openai/gpt-5-2025-08-07.toml | 24 -- .../models/openai/gpt-5-mini-2025-08-07.toml | 24 -- .../aimlapi/models/openai/gpt-5-mini.toml | 24 +- .../models/openai/gpt-5-nano-2025-08-07.toml | 24 -- .../aimlapi/models/openai/gpt-5-nano.toml | 24 +- .../aimlapi/models/openai/gpt-5-pro.toml | 24 +- .../models/openai/gpt-5.1-2025-11-13.toml | 24 -- .../models/openai/gpt-5.1-codex-max.toml | 10 + .../models/openai/gpt-5.1-codex-mini.toml | 24 +- .../aimlapi/models/openai/gpt-5.1-codex.toml | 24 +- .../models/openai/gpt-5.2-2025-12-11.toml | 24 -- .../models/openai/gpt-5.2-chat-latest.toml | 24 -- .../aimlapi/models/openai/gpt-5.2-codex.toml | 24 +- .../aimlapi/models/openai/gpt-5.2-pro.toml | 24 +- .../aimlapi/models/openai/gpt-5.3-codex.toml | 24 +- .../aimlapi/models/openai/gpt-5.4-mini.toml | 10 + .../aimlapi/models/openai/gpt-5.4-nano.toml | 10 + .../aimlapi/models/openai/gpt-5.4-pro.toml | 9 + .../aimlapi/models/openai/gpt-5.5-pro.toml | 9 + .../models/openai/gpt-5.6-luna-pro.toml | 10 + .../aimlapi/models/openai/gpt-5.6-luna.toml | 10 + .../models/openai/gpt-5.6-sol-pro.toml | 10 + .../aimlapi/models/openai/gpt-5.6-sol.toml | 10 + .../models/openai/gpt-5.6-terra-pro.toml | 10 + .../aimlapi/models/openai/gpt-5.6-terra.toml | 10 + providers/aimlapi/models/openai/gpt-5.toml | 24 +- .../aimlapi/models/openai/gpt-oss-120b.toml | 20 +- .../aimlapi/models/openai/gpt-oss-20b.toml | 20 +- providers/aimlapi/models/openai/o1-pro.toml | 9 + providers/aimlapi/models/openai/o1.toml | 13 + .../aimlapi/models/openai/o3-2025-04-16.toml | 24 -- providers/aimlapi/models/openai/o3-mini.toml | 10 + providers/aimlapi/models/openai/o3-pro.toml | 9 + .../models/openai/o4-mini-2025-04-16.toml | 24 -- .../aimlapi/models/perplexity/sonar-pro.toml | 17 +- .../aimlapi/models/perplexity/sonar.toml | 17 +- .../aimlapi/models/poolside/laguna-s-2.1.toml | 13 + .../models/poolside/laguna-xs-2.1.toml | 10 + .../aimlapi/models/sakana/fugu-ultra.toml | 21 +- .../aimlapi/models/sakana/sakana-namazu.toml | 24 +- .../models/stepfun/step-3.7-flash.toml | 24 -- .../aimlapi/models/tencent/hy-mt2-1.8b.toml | 23 -- .../models/tencent/hy-mt2-30b-a3b.toml | 23 -- .../aimlapi/models/tencent/hy-mt2-7b.toml | 23 -- providers/aimlapi/models/tencent/hy3.toml | 14 + .../aimlapi/models/tencent/hy4-preview.toml | 20 +- .../thinkingmachines/inkling-small.toml | 19 +- .../models/thinkingmachines/inkling.toml | 10 + .../aimlapi/models/upstage/solar-pro4.toml | 24 +- .../aimlapi/models/x-ai/grok-3-beta.toml | 24 -- .../aimlapi/models/x-ai/grok-3-mini-beta.toml | 24 -- .../x-ai/grok-4-1-fast-non-reasoning.toml | 24 -- .../models/x-ai/grok-4-1-fast-reasoning.toml | 24 -- .../x-ai/grok-4-20-0309-non-reasoning.toml | 24 -- .../models/x-ai/grok-4-20-0309-reasoning.toml | 24 -- providers/aimlapi/models/x-ai/grok-4-3.toml | 24 -- .../x-ai/grok-4-fast-non-reasoning.toml | 24 -- .../models/x-ai/grok-4-fast-reasoning.toml | 24 -- .../aimlapi/models/x-ai/grok-code-fast-1.toml | 24 -- .../aimlapi/models/xiaomi/mimo-v2.5-pro.toml | 20 +- .../aimlapi/models/xiaomi/mimo-v2.5.toml | 20 +- providers/aimlapi/models/z-ai/glm-4.5v.toml | 13 + providers/aimlapi/models/z-ai/glm-4.6v.toml | 13 + .../aimlapi/models/z-ai/glm-4.7-flash.toml | 14 + .../aimlapi/models/z-ai/glm-5.3-flash.toml | 24 -- .../aimlapi/models/zhipu/glm-4.5-air.toml | 24 -- providers/aimlapi/models/zhipu/glm-4.5.toml | 24 -- providers/aimlapi/models/zhipu/glm-4.6.toml | 24 -- providers/aimlapi/models/zhipu/glm-4.7.toml | 24 -- .../models/zhipu/glm-5-2-fast-preview.toml | 24 -- providers/aimlapi/models/zhipu/glm-5-2.toml | 24 -- providers/aimlapi/models/zhipu/glm-5-3.toml | 24 -- providers/aimlapi/models/zhipu/glm-5.2.toml | 24 -- providers/aimlapi/models/zhipu/glm-5.3.toml | 24 -- providers/aimlapi/models/zhipu/glm-5.toml | 24 -- 220 files changed, 984 insertions(+), 3638 deletions(-) delete mode 100644 providers/aimlapi/models/Qwen/Qwen2.5-7B-Instruct-Turbo.toml delete mode 100644 providers/aimlapi/models/Qwen/Qwen3-235B-A22B-Thinking-2507.toml delete mode 100644 providers/aimlapi/models/Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8.toml delete mode 100644 providers/aimlapi/models/alibaba/glm-5.2-fast-preview.toml delete mode 100644 providers/aimlapi/models/alibaba/glm-5.2.toml delete mode 100644 providers/aimlapi/models/alibaba/qwen-plus.toml delete mode 100644 providers/aimlapi/models/alibaba/qwen-turbo.toml delete mode 100644 providers/aimlapi/models/alibaba/qwen3-235b-a22b-thinking-2507.toml delete mode 100644 providers/aimlapi/models/alibaba/qwen3-32b.toml delete mode 100644 providers/aimlapi/models/alibaba/qwen3-max-preview.toml delete mode 100644 providers/aimlapi/models/alibaba/qwen3-next-80b-a3b-thinking.toml delete mode 100644 providers/aimlapi/models/alibaba/qwen3-omni-30b-a3b-captioner.toml delete mode 100644 providers/aimlapi/models/alibaba/qwen3-vl-32b-instruct.toml delete mode 100644 providers/aimlapi/models/alibaba/qwen3-vl-32b-thinking.toml delete mode 100644 providers/aimlapi/models/alibaba/qwen3-vl-flash.toml delete mode 100644 providers/aimlapi/models/alibaba/qwen3-vl-plus.toml delete mode 100644 providers/aimlapi/models/alibaba/qwen3.5-plus-20260218.toml delete mode 100644 providers/aimlapi/models/alibaba/qwen3.5-plus.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3.6-27b.toml create mode 100644 providers/aimlapi/models/alibaba/qwen3.6-35b-a3b.toml delete mode 100644 providers/aimlapi/models/alibaba/qwen3.6-flash.toml delete mode 100644 providers/aimlapi/models/alibaba/qwen3.6-max-preview.toml delete mode 100644 providers/aimlapi/models/alibaba/qwen3.6-plus.toml delete mode 100644 providers/aimlapi/models/alibaba/qwen3.7-plus.toml delete mode 100644 providers/aimlapi/models/anthracite-org/magnum-v4-72b.toml delete mode 100644 providers/aimlapi/models/anthropic/claude-fable-5.toml delete mode 100644 providers/aimlapi/models/anthropic/claude-haiku-4-5-20251001.toml delete mode 100644 providers/aimlapi/models/anthropic/claude-opus-4-1-20250805.toml delete mode 100644 providers/aimlapi/models/anthropic/claude-opus-4-5-20251101.toml delete mode 100644 providers/aimlapi/models/anthropic/claude-opus-4-6.toml delete mode 100644 providers/aimlapi/models/anthropic/claude-opus-4-8.toml create mode 100644 providers/aimlapi/models/anthropic/claude-opus-4.7-fast.toml create mode 100644 providers/aimlapi/models/anthropic/claude-opus-4.7.toml create mode 100644 providers/aimlapi/models/anthropic/claude-opus-4.8-fast.toml delete mode 100644 providers/aimlapi/models/anthropic/claude-opus-5.toml delete mode 100644 providers/aimlapi/models/anthropic/claude-sonnet-4-5-20250929.toml delete mode 100644 providers/aimlapi/models/anthropic/claude-sonnet-4-6.toml delete mode 100644 providers/aimlapi/models/anthropic/claude-sonnet-5.toml create mode 100644 providers/aimlapi/models/arcee-ai/trinity-large-thinking.toml delete mode 100644 providers/aimlapi/models/baidu/ernie-4.5-vl-424b-a47b.toml delete mode 100644 providers/aimlapi/models/baidu/ernie-5.0.toml delete mode 100644 providers/aimlapi/models/bytedance-seed/seed-2-1-turbo.toml create mode 100644 providers/aimlapi/models/bytedance-seed/seed-2.0-lite.toml create mode 100644 providers/aimlapi/models/bytedance-seed/seed-2.0-mini.toml delete mode 100644 providers/aimlapi/models/bytedance/seed-1-8.toml delete mode 100644 providers/aimlapi/models/cohere/command-a.toml create mode 100644 providers/aimlapi/models/cohere/command-r-08-2024.toml create mode 100644 providers/aimlapi/models/cohere/command-r-plus-08-2024.toml create mode 100644 providers/aimlapi/models/cohere/command-r7b-12-2024.toml delete mode 100644 providers/aimlapi/models/deepseek/deepseek-chat-v3.1.toml delete mode 100644 providers/aimlapi/models/deepseek/deepseek-non-reasoner-v3.1-terminus.toml delete mode 100644 providers/aimlapi/models/deepseek/deepseek-non-thinking-v3.2-exp.toml delete mode 100644 providers/aimlapi/models/deepseek/deepseek-r1.toml delete mode 100644 providers/aimlapi/models/deepseek/deepseek-reasoner-v3.1-terminus.toml delete mode 100644 providers/aimlapi/models/deepseek/deepseek-reasoner.toml delete mode 100644 providers/aimlapi/models/deepseek/deepseek-thinking-v3.2-exp.toml create mode 100644 providers/aimlapi/models/google/gemini-2.5-flash-image.toml delete mode 100644 providers/aimlapi/models/google/gemini-2.5-flash-lite-preview.toml delete mode 100644 providers/aimlapi/models/google/gemini-2.5-pro.toml delete mode 100644 providers/aimlapi/models/google/gemini-3-6-flash.toml delete mode 100644 providers/aimlapi/models/google/gemini-3-7-flash.toml delete mode 100644 providers/aimlapi/models/google/gemini-3-flash-preview.toml create mode 100644 providers/aimlapi/models/google/gemini-3-pro-image-preview.toml create mode 100644 providers/aimlapi/models/google/gemini-3-pro-image.toml create mode 100644 providers/aimlapi/models/google/gemini-3.1-flash-image-preview.toml create mode 100644 providers/aimlapi/models/google/gemini-3.1-flash-image.toml create mode 100644 providers/aimlapi/models/google/gemini-3.1-flash-lite-image.toml create mode 100644 providers/aimlapi/models/google/gemini-3.1-flash-lite-preview.toml create mode 100644 providers/aimlapi/models/google/gemini-3.1-pro-preview-customtools.toml delete mode 100644 providers/aimlapi/models/google/gemini-3.1-pro-preview.toml create mode 100644 providers/aimlapi/models/google/gemini-flash-latest.toml delete mode 100644 providers/aimlapi/models/google/gemma-3-12b-it.toml delete mode 100644 providers/aimlapi/models/google/gemma-3-27b-it.toml delete mode 100644 providers/aimlapi/models/google/gemma-3-4b-it.toml delete mode 100644 providers/aimlapi/models/google/gemma-4-26b-a4b-it-maas.toml create mode 100644 providers/aimlapi/models/google/gemma-4-31b-it.toml delete mode 100644 providers/aimlapi/models/gryphe/mythomax-l2-13b.toml delete mode 100644 providers/aimlapi/models/meituan/longcat-2.0.toml delete mode 100644 providers/aimlapi/models/meta-llama/Llama-3.3-70B-Instruct-Turbo.toml delete mode 100644 providers/aimlapi/models/meta-llama/llama-3.3-70b-versatile.toml create mode 100644 providers/aimlapi/models/meta/muse-glimmer-30b.toml create mode 100644 providers/aimlapi/models/meta/muse-spark-1.1.toml create mode 100644 providers/aimlapi/models/meta/muse-spark-1.2.toml delete mode 100644 providers/aimlapi/models/minimax/MiniMax-Text-01.toml delete mode 100644 providers/aimlapi/models/minimax/m1.toml delete mode 100644 providers/aimlapi/models/minimax/m2-1.toml delete mode 100644 providers/aimlapi/models/minimax/m2-5-20260218.toml delete mode 100644 providers/aimlapi/models/minimax/m2-5-highspeed-20260218.toml delete mode 100644 providers/aimlapi/models/minimax/m2.toml delete mode 100644 providers/aimlapi/models/minimax/minimax-m3.toml create mode 100644 providers/aimlapi/models/mistralai/devstral-2512.toml create mode 100644 providers/aimlapi/models/mistralai/mistral-large-2512.toml create mode 100644 providers/aimlapi/models/mistralai/mistral-small-2603.toml delete mode 100644 providers/aimlapi/models/moonshot/kimi-k2-5.toml delete mode 100644 providers/aimlapi/models/moonshot/kimi-k2-7-code-highspeed.toml delete mode 100644 providers/aimlapi/models/moonshot/kimi-k2-7-code.toml create mode 100644 providers/aimlapi/models/moonshotai/kimi-k2-thinking.toml delete mode 100644 providers/aimlapi/models/nousresearch/hermes-4-405b.toml delete mode 100644 providers/aimlapi/models/nvidia/nemotron-3-nano-30b-a3b.toml delete mode 100644 providers/aimlapi/models/nvidia/nemotron-3-super-120b-a12b.toml delete mode 100644 providers/aimlapi/models/nvidia/nemotron-3-ultra-550b-a55b.toml delete mode 100644 providers/aimlapi/models/nvidia/nemotron-3.5-lightning.toml create mode 100644 providers/aimlapi/models/openai/gpt-3.5-turbo.toml create mode 100644 providers/aimlapi/models/openai/gpt-4-turbo.toml delete mode 100644 providers/aimlapi/models/openai/gpt-4.1-2025-04-14.toml create mode 100644 providers/aimlapi/models/openai/gpt-4.1-mini.toml create mode 100644 providers/aimlapi/models/openai/gpt-4.1-nano.toml create mode 100644 providers/aimlapi/models/openai/gpt-4.toml create mode 100644 providers/aimlapi/models/openai/gpt-4o-mini.toml delete mode 100644 providers/aimlapi/models/openai/gpt-5-1.toml delete mode 100644 providers/aimlapi/models/openai/gpt-5-2-chat-latest.toml delete mode 100644 providers/aimlapi/models/openai/gpt-5-2.toml delete mode 100644 providers/aimlapi/models/openai/gpt-5-2025-08-07.toml delete mode 100644 providers/aimlapi/models/openai/gpt-5-mini-2025-08-07.toml delete mode 100644 providers/aimlapi/models/openai/gpt-5-nano-2025-08-07.toml delete mode 100644 providers/aimlapi/models/openai/gpt-5.1-2025-11-13.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.1-codex-max.toml delete mode 100644 providers/aimlapi/models/openai/gpt-5.2-2025-12-11.toml delete mode 100644 providers/aimlapi/models/openai/gpt-5.2-chat-latest.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.4-mini.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.4-nano.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.4-pro.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.5-pro.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.6-luna-pro.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.6-luna.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.6-sol-pro.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.6-sol.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.6-terra-pro.toml create mode 100644 providers/aimlapi/models/openai/gpt-5.6-terra.toml create mode 100644 providers/aimlapi/models/openai/o1-pro.toml create mode 100644 providers/aimlapi/models/openai/o1.toml delete mode 100644 providers/aimlapi/models/openai/o3-2025-04-16.toml create mode 100644 providers/aimlapi/models/openai/o3-mini.toml create mode 100644 providers/aimlapi/models/openai/o3-pro.toml delete mode 100644 providers/aimlapi/models/openai/o4-mini-2025-04-16.toml create mode 100644 providers/aimlapi/models/poolside/laguna-s-2.1.toml create mode 100644 providers/aimlapi/models/poolside/laguna-xs-2.1.toml delete mode 100644 providers/aimlapi/models/stepfun/step-3.7-flash.toml delete mode 100644 providers/aimlapi/models/tencent/hy-mt2-1.8b.toml delete mode 100644 providers/aimlapi/models/tencent/hy-mt2-30b-a3b.toml delete mode 100644 providers/aimlapi/models/tencent/hy-mt2-7b.toml create mode 100644 providers/aimlapi/models/tencent/hy3.toml create mode 100644 providers/aimlapi/models/thinkingmachines/inkling.toml delete mode 100644 providers/aimlapi/models/x-ai/grok-3-beta.toml delete mode 100644 providers/aimlapi/models/x-ai/grok-3-mini-beta.toml delete mode 100644 providers/aimlapi/models/x-ai/grok-4-1-fast-non-reasoning.toml delete mode 100644 providers/aimlapi/models/x-ai/grok-4-1-fast-reasoning.toml delete mode 100644 providers/aimlapi/models/x-ai/grok-4-20-0309-non-reasoning.toml delete mode 100644 providers/aimlapi/models/x-ai/grok-4-20-0309-reasoning.toml delete mode 100644 providers/aimlapi/models/x-ai/grok-4-3.toml delete mode 100644 providers/aimlapi/models/x-ai/grok-4-fast-non-reasoning.toml delete mode 100644 providers/aimlapi/models/x-ai/grok-4-fast-reasoning.toml delete mode 100644 providers/aimlapi/models/x-ai/grok-code-fast-1.toml create mode 100644 providers/aimlapi/models/z-ai/glm-4.5v.toml create mode 100644 providers/aimlapi/models/z-ai/glm-4.6v.toml create mode 100644 providers/aimlapi/models/z-ai/glm-4.7-flash.toml delete mode 100644 providers/aimlapi/models/z-ai/glm-5.3-flash.toml delete mode 100644 providers/aimlapi/models/zhipu/glm-4.5-air.toml delete mode 100644 providers/aimlapi/models/zhipu/glm-4.5.toml delete mode 100644 providers/aimlapi/models/zhipu/glm-4.6.toml delete mode 100644 providers/aimlapi/models/zhipu/glm-4.7.toml delete mode 100644 providers/aimlapi/models/zhipu/glm-5-2-fast-preview.toml delete mode 100644 providers/aimlapi/models/zhipu/glm-5-2.toml delete mode 100644 providers/aimlapi/models/zhipu/glm-5-3.toml delete mode 100644 providers/aimlapi/models/zhipu/glm-5.2.toml delete mode 100644 providers/aimlapi/models/zhipu/glm-5.3.toml delete mode 100644 providers/aimlapi/models/zhipu/glm-5.toml diff --git a/packages/core/src/sync/providers/aimlapi.ts b/packages/core/src/sync/providers/aimlapi.ts index ebfb255fa26..c575a607f18 100755 --- a/packages/core/src/sync/providers/aimlapi.ts +++ b/packages/core/src/sync/providers/aimlapi.ts @@ -1,49 +1,72 @@ import { z } from "zod"; -import { inferKimiFamily, ModelFamilyValues } from "../../family.js"; import type { SyncProvider } from "../index.js"; +import { factorBaseModel, modelMetadata, resolveModelMetadataBaseModel } from "./openrouter.js"; // The public catalog needs no key, and `include` is what turns on the pricing // and modality blocks this sync depends on. const API_ENDPOINT = "https://api.aimlapi.com/v1/models?include=pricing,modalities"; +// Per-model request schema. It is the only place the API states which reasoning +// controls a model actually accepts, so reasoning_options is read from here +// rather than assumed. +const DOCS_ENDPOINT = "https://api.aimlapi.com/docs-json"; + // AI/ML API serves one id under several endpoint types — a model can be both a // chat model and, say, an image model. Only the chat surface belongs here. const CHAT_COMPLETIONS_TYPE = "openai/chat-completions"; -const PricingUnit = z.object({ - name: z.string().nullish(), - content: z.string().nullish(), - origin: z.string().nullish(), - price: z.number().nullish(), - per: z.number().nullish(), -}).passthrough(); - -const Info = z.object({ - name: z.string().nullish(), - description: z.string().nullish(), - developer: z.string().nullish(), - releasedAt: z.string().nullish(), - contextLength: z.number().int().nonnegative().nullish(), - outputMax: z.number().int().nonnegative().nullish(), -}).passthrough(); - -export const AimlapiModel = z.object({ - id: z.string().min(1), - type: z.string().nullish(), - info: Info.nullish(), - modalities: z.object({ - input: z.array(z.string()).nullish(), - output: z.array(z.string()).nullish(), - }).passthrough().nullish(), - pricing: z.object({ - units: z.array(PricingUnit).nullish(), - }).passthrough().nullish(), -}).passthrough(); - -export const AimlapiResponse = z.object({ - data: z.array(AimlapiModel).min(1), -}).passthrough(); +// Values this schema accepts for an "effort" reasoning control. Anything the +// API documents outside this set is dropped rather than coerced. +const EFFORT_VALUES = new Set(["none", "minimal", "low", "medium", "high", "xhigh", "max", "default"]); + +const DOCS_CONCURRENCY = 8; + +const PricingUnit = z + .object({ + name: z.string().nullish(), + content: z.string().nullish(), + origin: z.string().nullish(), + price: z.number().nullish(), + per: z.number().nullish(), + }) + .passthrough(); + +const Info = z + .object({ + contextLength: z.number().int().nonnegative().nullish(), + outputMax: z.number().int().nonnegative().nullish(), + }) + .passthrough(); + +export const AimlapiModel = z + .object({ + id: z.string().min(1), + type: z.string().nullish(), + info: Info.nullish(), + modalities: z + .object({ + input: z.array(z.string()).nullish(), + output: z.array(z.string()).nullish(), + }) + .passthrough() + .nullish(), + pricing: z + .object({ + units: z.array(PricingUnit).nullish(), + }) + .passthrough() + .nullish(), + /** Attached by fetchModels; not part of the upstream payload. */ + reasoningEffort: z.array(z.string()).nullish(), + }) + .passthrough(); + +export const AimlapiResponse = z + .object({ + data: z.array(AimlapiModel).min(1), + }) + .passthrough(); export type AimlapiModel = z.infer; @@ -61,19 +84,40 @@ function normalizeModalities(values: readonly string[] | null | undefined): Moda return [...seen]; } +function isChatTextModel(model: AimlapiModel): boolean { + if (model.type !== CHAT_COMPLETIONS_TYPE) return false; + const output = normalizeModalities(model.modalities?.output); + // A chat model whose output is not purely text is a media model riding the + // chat protocol, and does not belong in a chat catalog. + return output.length === 1 && output[0] === "text"; +} + +/** + * Lab entry this id is a host for. AI/ML API is an aggregator and authors none + * of these models, so every entry has to point at the lab file rather than + * restate it. + */ +function baseModelFor(id: string): string | undefined { + return resolveModelMetadataBaseModel(id); +} + +function baseReasoning(baseModelID: string): boolean { + try { + return modelMetadata(baseModelID).reasoning === true; + } catch { + return false; + } +} + /** * Prices are quoted as `price` per `per` tokens; models.dev stores dollars per * million. The unit discriminator is `origin`, not `measure`: provided is * input, generated is output, cached is a cache read. Only text token charges * are taken — a model's image or audio units are a different surface. */ -function perMillion( - units: readonly z.infer[], - origin: string, -): number | undefined { +function perMillion(units: readonly z.infer[], origin: string): number | undefined { const unit = units.find( - (candidate) => - candidate.name === "token" && candidate.content === "text" && candidate.origin === origin, + (candidate) => candidate.name === "token" && candidate.content === "text" && candidate.origin === origin, ); if (!unit || unit.price == null || !unit.per) return undefined; return (unit.price / unit.per) * 1_000_000; @@ -84,31 +128,72 @@ function positive(value: number | null | undefined): number | undefined { } /** - * `family` is a closed vocabulary of model families (`gpt`, `claude`, `o`, …), - * not the vendor prefix — matching on the prefix produces values the schema - * rejects. Longest match first so `claude-sonnet` wins over `claude`, and `o` - * only matches when a digit follows it, as in `o3`. + * Reads the documented `reasoning_effort` enum for one model. Returns undefined + * when the docs do not describe the control, which is treated as "cannot state + * it" rather than "the model has none". */ -function inferFamily(id: string, name: string): string | undefined { - const kimi = inferKimiFamily(id, name); - if (kimi !== undefined) return kimi; - - const target = `${id} ${name}`.toLowerCase(); - return [...ModelFamilyValues] - .sort((a, b) => b.length - a.length) - .find((family) => { - const value = family.toLowerCase().replace(/[.*+?^${}()|[\]\\]/g, "\\$&"); - if (family === "o") { - return new RegExp(`(^|[^a-z0-9])${value}(?=\\d|$|[^a-z0-9])`).test(target); - } - return new RegExp(`(^|[^a-z0-9])${value}(?=$|[^a-z0-9])`).test(target); - }); +async function fetchReasoningEffort(id: string): Promise { + const url = `${DOCS_ENDPOINT}?model=${encodeURIComponent(id)}&endpoint=${encodeURIComponent(CHAT_COMPLETIONS_TYPE)}`; + let payload: unknown; + try { + const response = await fetch(url); + if (!response.ok) return undefined; + payload = await response.json(); + } catch { + return undefined; + } + + const found = findReasoningEffortEnum(payload); + if (found === undefined) return undefined; + + const values = found.filter((value) => EFFORT_VALUES.has(value)); + return values.length > 0 ? values : undefined; } -function releaseDate(value: string | null | undefined): string | undefined { - if (!value) return undefined; - const match = /^(\d{4}-\d{2}-\d{2})/.exec(value.trim()); - return match?.[1]; +function findReasoningEffortEnum(node: unknown): string[] | undefined { + if (Array.isArray(node)) { + for (const item of node) { + const found = findReasoningEffortEnum(item); + if (found !== undefined) return found; + } + return undefined; + } + if (node === null || typeof node !== "object") return undefined; + + const record = node as Record; + const effort = record["reasoning_effort"]; + if (effort !== null && typeof effort === "object") { + const values = (effort as Record)["enum"]; + if (Array.isArray(values) && values.every((value) => typeof value === "string")) { + return values as string[]; + } + } + + for (const value of Object.values(record)) { + const found = findReasoningEffortEnum(value); + if (found !== undefined) return found; + } + return undefined; +} + +async function attachReasoningEffort(models: AimlapiModel[]): Promise { + // Only models whose lab entry says they reason need the control documented, + // and only those are worth a request. + const pending = models.filter((model) => { + if (!isChatTextModel(model)) return false; + const base = baseModelFor(model.id); + return base !== undefined && baseReasoning(base); + }); + + let cursor = 0; + const workers = Array.from({ length: Math.min(DOCS_CONCURRENCY, pending.length) }, async () => { + while (cursor < pending.length) { + const model = pending[cursor++]; + if (model === undefined) return; + model.reasoningEffort = await fetchReasoningEffort(model.id); + } + }); + await Promise.all(workers); } export const aimlapi = { @@ -119,12 +204,12 @@ export const aimlapi = { // a local model missing from one response is not proof that it is gone. deleteMissing: false, sourceID(model) { - return model.type === CHAT_COMPLETIONS_TYPE ? model.id : undefined; + return isChatTextModel(model) ? model.id : undefined; }, skippedNotice(ids) { if (ids.length === 0) return []; return [ - `${ids.length} AI/ML API chat models were skipped because the catalog does not yet publish a description or an output limit for them, both of which this schema requires.`, + `${ids.length} AI/ML API chat models were skipped because this repository has no lab entry to point \`base_model\` at, or because the API does not document the reasoning control a reasoning model requires.`, `Skipped remote IDs: ${ids.map((id) => `\`${id}\``).join(", ")}`, ]; }, @@ -140,67 +225,69 @@ export const aimlapi = { if (!response.ok) { throw new Error(`AI/ML API request failed: ${response.status} ${response.statusText}`); } - return response.json(); + const raw = await response.json(); + const parsed = AimlapiResponse.parse(raw); + await attachReasoningEffort(parsed.data); + return parsed; }, parseModels(raw) { return AimlapiResponse.parse(raw).data; }, translateModel(model, context) { - // Chat only. The same id may also appear under an image or audio type; - // those entries are a different product and are skipped silently. - if (model.type !== CHAT_COMPLETIONS_TYPE) return undefined; + if (!isChatTextModel(model)) return undefined; const existing = context.existing(model.id); - const info = model.info ?? {}; - const units = model.pricing?.units ?? []; - const output = normalizeModalities(model.modalities?.output); - // A chat model whose output is not purely text is a media model riding the - // chat protocol, and does not belong in a chat catalog. - if (output.length !== 1 || output[0] !== "text") return undefined; + // AI/ML API hosts other people's models, so the entry must reference the + // lab file instead of duplicating it. Without a lab entry to point at there + // is nothing correct to write: inlining the metadata is what this schema + // forbids, and authoring the lab file would mean sourcing capability data + // the catalog does not publish. + const base = existing?.base_model ?? baseModelFor(model.id); + if (base === undefined) return undefined; - const input = normalizeModalities(model.modalities?.input); - const contextLimit = positive(info.contextLength) ?? existing?.limit?.context; - const outputLimit = positive(info.outputMax) ?? existing?.limit?.output; - const description = info.description?.trim() || existing?.description; + // Required whenever the base model reasons. Only the API's own request + // schema can say which values it takes, so a model whose docs stay silent + // is skipped rather than given an invented control. + let reasoningOptions: Array<{ type: "effort"; values: string[] }> | undefined; + if (baseReasoning(base)) { + const values = model.reasoningEffort ?? undefined; + if (values === undefined || values.length === 0) return undefined; + reasoningOptions = [{ type: "effort", values }]; + } - // The schema requires a description and an output limit. Where the catalog - // publishes neither and no local value exists, the model is skipped rather - // than filled in: an invented description or a guessed limit would be worse - // than an absent entry, and skippedNotice makes the gap visible. - if (!description || outputLimit === undefined) return undefined; + const units = model.pricing?.units ?? []; + const info = model.info ?? {}; + const contextLimit = positive(info.contextLength); + const outputLimit = positive(info.outputMax); + // Only what the catalog actually publishes. It reports a context window and + // an output cap but no input cap, and equating the input cap with the whole + // context would overwrite the lab's correct split (e.g. 272k in + 128k out + // within a 400k window) with a wrong number. + const limit = + contextLimit === undefined && outputLimit === undefined + ? undefined + : { context: contextLimit, output: outputLimit }; + // Everything else — the capability flags, description, dates, modalities — + // is the lab's to state and is inherited. factorBaseModel drops whatever + // matches the base, so the file carries only what is genuinely ours. return { id: model.id, - model: { - name: info.name?.trim() || existing?.name || model.id, - description, - family: existing?.family ?? inferFamily(model.id, info.name?.trim() ?? ""), - release_date: releaseDate(info.releasedAt) ?? existing?.release_date, - last_updated: releaseDate(info.releasedAt) ?? existing?.last_updated, - // The catalog does not report these capabilities. A value already in the - // repo was put there by someone who checked; a default here would only - // overwrite that with a guess. - attachment: existing?.attachment ?? input.length > 1, - reasoning: existing?.reasoning ?? false, - tool_call: existing?.tool_call ?? false, - structured_output: existing?.structured_output ?? false, - open_weights: existing?.open_weights ?? false, - cost: { - input: perMillion(units, "provided") ?? existing?.cost?.input, - output: perMillion(units, "generated") ?? existing?.cost?.output, - cache_read: perMillion(units, "cached") ?? existing?.cost?.cache_read, - }, - limit: { - context: contextLimit, - input: contextLimit, - output: outputLimit, - }, - modalities: { - input, - output, + model: factorBaseModel( + base, + { + cost: { + input: perMillion(units, "provided") ?? existing?.cost?.input, + output: perMillion(units, "generated") ?? existing?.cost?.output, + cache_read: perMillion(units, "cached") ?? existing?.cost?.cache_read, + }, + reasoning_options: reasoningOptions, + limit, }, - }, + limit, + existing?.base_model === base ? existing?.base_model_omit : undefined, + ), }; }, } satisfies SyncProvider; diff --git a/providers/aimlapi/models/Qwen/Qwen2.5-7B-Instruct-Turbo.toml b/providers/aimlapi/models/Qwen/Qwen2.5-7B-Instruct-Turbo.toml deleted file mode 100644 index 158a428d08e..00000000000 --- a/providers/aimlapi/models/Qwen/Qwen2.5-7B-Instruct-Turbo.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Qwen2.5 7B Instruct Turbo" -description = "Qwen 2.5 7B Instruct Turbo excels in coding and instruction following." -family = "qwen" -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.39 -output = 0.39 - -[limit] -context = 32_000 -input = 32_000 -output = 31_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/Qwen/Qwen3-235B-A22B-Thinking-2507.toml b/providers/aimlapi/models/Qwen/Qwen3-235B-A22B-Thinking-2507.toml deleted file mode 100644 index 4d8c764a966..00000000000 --- a/providers/aimlapi/models/Qwen/Qwen3-235B-A22B-Thinking-2507.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Qwen3 Thinking 2507" -description = "Alibaba's Qwen 3 thinking 2507 is a multilingual reasoning model with knowledge augmentation, and creative capabilities." -family = "qwen" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.845 -output = 3.9 - -[limit] -context = 32_000 -input = 32_000 -output = 16_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8.toml b/providers/aimlapi/models/Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8.toml deleted file mode 100644 index dc0491c81f8..00000000000 --- a/providers/aimlapi/models/Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Qwen3 Coder" -description = "Alibaba's Qwen 3 Coder is a powerful coding agent." -family = "qwen" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 2.6 -output = 2.6 - -[limit] -context = 262_000 -input = 262_000 -output = 65_536 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/glm-5.2-fast-preview.toml b/providers/aimlapi/models/alibaba/glm-5.2-fast-preview.toml deleted file mode 100644 index e6cf4b4c3bd..00000000000 --- a/providers/aimlapi/models/alibaba/glm-5.2-fast-preview.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GLM 5.2 Fast Preview" -description = "A latency-optimized GLM-5.2 preview with the same 1M-token context window and 1.5–2× higher output throughput for real-time chat, agents, and coding." -family = "glm" -release_date = "2026-07-10" -last_updated = "2026-07-10" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 4.55 -output = 14.3 -cache_read = 0.91 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 131_072 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/glm-5.2.toml b/providers/aimlapi/models/alibaba/glm-5.2.toml deleted file mode 100644 index f012ea0d537..00000000000 --- a/providers/aimlapi/models/alibaba/glm-5.2.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GLM 5.2" -description = "GLM-5.2 is Zhipu AI flagship MoE LLM with a 1M-token context window, optimized for agentic coding, tool use, reasoning, and long-context tasks." -family = "glm" -release_date = "2026-06-16" -last_updated = "2026-06-16" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.82 -output = 5.72 -cache_read = 0.364 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 131_072 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen-max.toml b/providers/aimlapi/models/alibaba/qwen-max.toml index 4554f9eba34..b67b2dfda47 100644 --- a/providers/aimlapi/models/alibaba/qwen-max.toml +++ b/providers/aimlapi/models/alibaba/qwen-max.toml @@ -1,13 +1,4 @@ -name = "Qwen Max" -description = "Qwen Max: powerful AI, excels in stability, competes globally." -family = "qwen" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "alibaba/qwen-max" [cost] input = 2.08 @@ -16,9 +7,3 @@ cache_read = 0.416 [limit] context = 32_000 -input = 32_000 -output = 8_192 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen-plus.toml b/providers/aimlapi/models/alibaba/qwen-plus.toml deleted file mode 100644 index 29e5653ccb5..00000000000 --- a/providers/aimlapi/models/alibaba/qwen-plus.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Qwen Plus" -description = "Qwen-Plus: Alibaba's advanced multilingual model with enhanced reasoning and instruction-following capabilities." -family = "qwen" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.52 -output = 1.56 -cache_read = 0.104 - -[limit] -context = 131_000 -input = 131_000 -output = 16_384 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen-turbo.toml b/providers/aimlapi/models/alibaba/qwen-turbo.toml deleted file mode 100644 index c1cd95d2ef1..00000000000 --- a/providers/aimlapi/models/alibaba/qwen-turbo.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Qwen Turbo" -description = "Qwen Turbo: optimizes AI agent speed, integrates with RAG, large context window." -family = "qwen" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.065 -output = 0.26 -cache_read = 0.013 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 16_384 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-235b-a22b-thinking-2507.toml b/providers/aimlapi/models/alibaba/qwen3-235b-a22b-thinking-2507.toml deleted file mode 100644 index 492d06a2607..00000000000 --- a/providers/aimlapi/models/alibaba/qwen3-235b-a22b-thinking-2507.toml +++ /dev/null @@ -1,22 +0,0 @@ -name = "Qwen3 Thinking 2507" -description = "Alibaba's Qwen 3 thinking 2507 is a multilingual reasoning model with knowledge augmentation, and creative capabilities." -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.299 -output = 2.99 - -[limit] -context = 32_000 -input = 32_000 -output = 16_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-32b.toml b/providers/aimlapi/models/alibaba/qwen3-32b.toml deleted file mode 100644 index d75ce76e131..00000000000 --- a/providers/aimlapi/models/alibaba/qwen3-32b.toml +++ /dev/null @@ -1,22 +0,0 @@ -name = "Qwen3 32B" -description = "Alibaba's Qwen 3 32B is a multilingual model with instruction-following capabilities and enhanced reasoning." -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.208 -output = 0.832 - -[limit] -context = 131_000 -input = 131_000 -output = 16_384 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-coder-480b-a35b-instruct.toml b/providers/aimlapi/models/alibaba/qwen3-coder-480b-a35b-instruct.toml index af1e1067450..c294f7401f9 100644 --- a/providers/aimlapi/models/alibaba/qwen3-coder-480b-a35b-instruct.toml +++ b/providers/aimlapi/models/alibaba/qwen3-coder-480b-a35b-instruct.toml @@ -1,12 +1,4 @@ -name = "Qwen3 Coder" -description = "Alibaba's Qwen 3 Coder is a powerful coding agent." -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "alibaba/qwen3-coder-480b-a35b-instruct" [cost] input = 1.95 @@ -14,9 +6,3 @@ output = 9.75 [limit] context = 262_000 -input = 262_000 -output = 65_536 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-max-preview.toml b/providers/aimlapi/models/alibaba/qwen3-max-preview.toml deleted file mode 100644 index f78f0fcb224..00000000000 --- a/providers/aimlapi/models/alibaba/qwen3-max-preview.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Qwen3 Max Preview" -description = "Alibaba's Qwen 3 Max Preview is a multilingual model with instruction-following capabilities and enhanced reasoning." -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.56 -output = 7.8 -cache_read = 0.312 - -[limit] -context = 252_000 -input = 252_000 -output = 32_768 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-max.toml b/providers/aimlapi/models/alibaba/qwen3-max.toml index 81be3f5718a..2cc25e4c73e 100644 --- a/providers/aimlapi/models/alibaba/qwen3-max.toml +++ b/providers/aimlapi/models/alibaba/qwen3-max.toml @@ -1,23 +1,6 @@ -name = "Qwen3 Max Instruct" -description = "Alibaba's Qwen 3 Max Instruct is the most powerful model in the Qwen 3 series." -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "alibaba/qwen3-max" [cost] input = 1.56 output = 7.8 cache_read = 0.312 - -[limit] -context = 262_144 -input = 262_144 -output = 65_536 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-next-80b-a3b-instruct.toml b/providers/aimlapi/models/alibaba/qwen3-next-80b-a3b-instruct.toml index 8081cee7649..b90216744f6 100644 --- a/providers/aimlapi/models/alibaba/qwen3-next-80b-a3b-instruct.toml +++ b/providers/aimlapi/models/alibaba/qwen3-next-80b-a3b-instruct.toml @@ -1,12 +1,4 @@ -name = "Qwen3 Next 80B A3B Instruct" -description = "Alibaba's Qwen 3 Next 80B A3B Instruct is a multilingual model with enhanced instruction-following capabilities." -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "alibaba/qwen3-next-80b-a3b-instruct" [cost] input = 0.195 @@ -14,9 +6,4 @@ output = 1.56 [limit] context = 129_024 -input = 129_024 output = 16_384 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-next-80b-a3b-thinking.toml b/providers/aimlapi/models/alibaba/qwen3-next-80b-a3b-thinking.toml deleted file mode 100644 index 5640e1c0adb..00000000000 --- a/providers/aimlapi/models/alibaba/qwen3-next-80b-a3b-thinking.toml +++ /dev/null @@ -1,22 +0,0 @@ -name = "Qwen3 Next 80B A3B Thinking" -description = "Alibaba's Qwen 3 Next 80B A3B Thinking is a multilingual reasoning model with advanced chain-of-thought capabilities." -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.195 -output = 1.56 - -[limit] -context = 126_976 -input = 126_976 -output = 81_920 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-omni-30b-a3b-captioner.toml b/providers/aimlapi/models/alibaba/qwen3-omni-30b-a3b-captioner.toml deleted file mode 100644 index 134b6d09af6..00000000000 --- a/providers/aimlapi/models/alibaba/qwen3-omni-30b-a3b-captioner.toml +++ /dev/null @@ -1,22 +0,0 @@ -name = "Qwen3 Omni Captioner" -description = "Qwen3-Omni-Captioner automatically generates accurate and comprehensive descriptions for complex audio, including speech, ambient sounds, music, and sound effects. The model can identify speaker emotions, musical elements, and is suitable for audio content analysis, security audits, and audio editing. Supports audio input up to 40 minutes (1 second = 25 tokens)." -release_date = "2025-11-14" -last_updated = "2025-11-14" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 4.953 -output = 3.9779999999999998 - -[limit] -context = 65_536 -input = 65_536 -output = 32_768 - -[modalities] -input = ["audio"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-vl-32b-instruct.toml b/providers/aimlapi/models/alibaba/qwen3-vl-32b-instruct.toml deleted file mode 100644 index cebfcf225b2..00000000000 --- a/providers/aimlapi/models/alibaba/qwen3-vl-32b-instruct.toml +++ /dev/null @@ -1,22 +0,0 @@ -name = "Qwen3 VL 32B Instruct" -description = "Alibaba's Qwen 3 VL 32B Instruct is a powerful vision-language model with hybrid instruction-following capabilities." -release_date = "2025-11-11" -last_updated = "2025-11-11" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.91 -output = 3.64 - -[limit] -context = 126_000 -input = 126_000 -output = 32_768 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-vl-32b-thinking.toml b/providers/aimlapi/models/alibaba/qwen3-vl-32b-thinking.toml deleted file mode 100644 index 40b6f7e4b14..00000000000 --- a/providers/aimlapi/models/alibaba/qwen3-vl-32b-thinking.toml +++ /dev/null @@ -1,22 +0,0 @@ -name = "Qwen3 VL 32B Thinking" -description = "Alibaba's Qwen 3 VL 32B Thinking is a powerful vision-language model with hybrid thinking capabilities." -release_date = "2025-11-11" -last_updated = "2025-11-11" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.91 -output = 10.92 - -[limit] -context = 126_000 -input = 126_000 -output = 32_768 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-vl-flash.toml b/providers/aimlapi/models/alibaba/qwen3-vl-flash.toml deleted file mode 100644 index bb3450bfa3f..00000000000 --- a/providers/aimlapi/models/alibaba/qwen3-vl-flash.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Qwen3 VL Flash" -description = "Alibaba's Qwen 3 VL Flash is a fast and cost-effective vision-language model with hybrid thinking capabilities. Supports image understanding with up to 16,384 tokens per image and high-resolution mode for detailed analysis." -release_date = "2025-10-19" -last_updated = "2025-10-19" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.065 -output = 0.52 -cache_read = 0.065 - -[limit] -context = 262_144 -input = 262_144 -output = 32_768 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3-vl-plus.toml b/providers/aimlapi/models/alibaba/qwen3-vl-plus.toml deleted file mode 100644 index cb26d7e6e87..00000000000 --- a/providers/aimlapi/models/alibaba/qwen3-vl-plus.toml +++ /dev/null @@ -1,22 +0,0 @@ -name = "Qwen3 VL Plus" -description = "Alibaba's Qwen 3 VL Plus is a powerful vision-language model with hybrid thinking capabilities. Supports image understanding with up to 16,384 tokens per image and high-resolution mode for detailed analysis." -release_date = "2025-10-19" -last_updated = "2025-10-19" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.26 -output = 2.08 - -[limit] -context = 262_144 -input = 262_144 -output = 32_768 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.5-flash.toml b/providers/aimlapi/models/alibaba/qwen3.5-flash.toml index 82c9f83d08f..ce03c2805d0 100644 --- a/providers/aimlapi/models/alibaba/qwen3.5-flash.toml +++ b/providers/aimlapi/models/alibaba/qwen3.5-flash.toml @@ -1,23 +1,9 @@ -name = "Qwen3.5 Flash" -description = "Qwen Max: powerful AI, excels in stability, competes globally." -family = "qwen3.5" -release_date = "2026-04-21" -last_updated = "2026-04-21" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "alibaba/qwen3.5-flash" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.13 output = 0.52 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 65_536 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.5-plus-20260218.toml b/providers/aimlapi/models/alibaba/qwen3.5-plus-20260218.toml deleted file mode 100644 index 523acb8c130..00000000000 --- a/providers/aimlapi/models/alibaba/qwen3.5-plus-20260218.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Qwen3.5 Plus" -description = "Qwen3.5-Plus is a commercial large language model by Alibaba Cloud designed for long-context text generation and enterprise-grade conversational AI. Supports up to 1M tokens per request with production-ready API stability." -family = "qwen3.5" -release_date = "2026-02-20" -last_updated = "2026-02-20" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.52 -output = 3.12 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 65_536 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.5-plus.toml b/providers/aimlapi/models/alibaba/qwen3.5-plus.toml deleted file mode 100644 index 523acb8c130..00000000000 --- a/providers/aimlapi/models/alibaba/qwen3.5-plus.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Qwen3.5 Plus" -description = "Qwen3.5-Plus is a commercial large language model by Alibaba Cloud designed for long-context text generation and enterprise-grade conversational AI. Supports up to 1M tokens per request with production-ready API stability." -family = "qwen3.5" -release_date = "2026-02-20" -last_updated = "2026-02-20" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.52 -output = 3.12 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 65_536 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.6-27b.toml b/providers/aimlapi/models/alibaba/qwen3.6-27b.toml new file mode 100644 index 00000000000..99848065e45 --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3.6-27b.toml @@ -0,0 +1,13 @@ +base_model = "alibaba/qwen3.6-27b" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.78 +output = 4.68 + +[limit] +context = 256_000 +output = 252_000 diff --git a/providers/aimlapi/models/alibaba/qwen3.6-35b-a3b.toml b/providers/aimlapi/models/alibaba/qwen3.6-35b-a3b.toml new file mode 100644 index 00000000000..71fb076430e --- /dev/null +++ b/providers/aimlapi/models/alibaba/qwen3.6-35b-a3b.toml @@ -0,0 +1,13 @@ +base_model = "alibaba/qwen3.6-35b-a3b" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.4875 +output = 2.925 + +[limit] +context = 256_000 +output = 252_000 diff --git a/providers/aimlapi/models/alibaba/qwen3.6-flash.toml b/providers/aimlapi/models/alibaba/qwen3.6-flash.toml deleted file mode 100644 index 7a18f32d8ce..00000000000 --- a/providers/aimlapi/models/alibaba/qwen3.6-flash.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Qwen3.6 Flash" -description = "Qwen3.6-Flash — a lightweight and cost-efficient model by Alibaba Cloud designed for simple tasks and high-throughput applications. It offers fast response times with lower cost." -family = "qwen3.6" -release_date = "2026-04-21" -last_updated = "2026-04-21" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.325 -output = 1.95 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 65_536 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.6-max-preview.toml b/providers/aimlapi/models/alibaba/qwen3.6-max-preview.toml deleted file mode 100644 index 50cc3406ff5..00000000000 --- a/providers/aimlapi/models/alibaba/qwen3.6-max-preview.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Qwen3.6 Max Preview" -description = "Qwen3.6-Max-Preview — a high-capability reasoning model by Alibaba Cloud designed for complex tasks and long-form problem solving. It provides strong reasoning performance and high accuracy for advanced use cases." -family = "qwen3.6" -release_date = "2026-04-21" -last_updated = "2026-04-21" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.65 -output = 3.9 - -[limit] -context = 262_144 -input = 262_144 -output = 65_536 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.6-plus.toml b/providers/aimlapi/models/alibaba/qwen3.6-plus.toml deleted file mode 100644 index df64485cf70..00000000000 --- a/providers/aimlapi/models/alibaba/qwen3.6-plus.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Qwen3.6 Plus" -description = "Qwen3.6-Plus — a balanced general-purpose model by Alibaba Cloud optimized for performance, cost, and speed. Suitable for a wide range of chat and generation tasks." -family = "qwen3.6" -release_date = "2026-04-21" -last_updated = "2026-04-21" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.52 -output = 3.12 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 65_536 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.7-max.toml b/providers/aimlapi/models/alibaba/qwen3.7-max.toml index e90f8474b8a..ad69c73c11c 100644 --- a/providers/aimlapi/models/alibaba/qwen3.7-max.toml +++ b/providers/aimlapi/models/alibaba/qwen3.7-max.toml @@ -1,24 +1,10 @@ -name = "Qwen3.7 Max" -description = "Qwen3.7-Max is Alibaba’s flagship reasoning and agentic LLM optimized for coding, productivity, long-context understanding and autonomous workflows. Text input and text output only. Supports function calling, explicit context cache and advanced reasoning capabilities." -family = "qwen3.7-max" -release_date = "2026-05-21" -last_updated = "2026-05-21" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "alibaba/qwen3.7-max" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 2.6 output = 7.8 cache_read = 2.6 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 65_536 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.7-plus.toml b/providers/aimlapi/models/alibaba/qwen3.7-plus.toml deleted file mode 100644 index 64173d1ebc5..00000000000 --- a/providers/aimlapi/models/alibaba/qwen3.7-plus.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Qwen3.7-Plus" -description = "Qwen3.7-Plus is a multimodal agent model that unifies vision and language into a single agent foundation. It delivers strong vision-language capabilities while retaining agentic strength in coding, tool use, and productivity workflows — including GUI understanding, visual Q&A, and full-modality input for complex software engineering tasks." -family = "qwen3.7-plus" -release_date = "2026-06-01" -last_updated = "2026-06-01" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.416 -output = 1.664 -cache_read = 0.0832 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 65_536 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.8-2.4t-a95b.toml b/providers/aimlapi/models/alibaba/qwen3.8-2.4t-a95b.toml index eedc3ee7545..3ae6076152a 100644 --- a/providers/aimlapi/models/alibaba/qwen3.8-2.4t-a95b.toml +++ b/providers/aimlapi/models/alibaba/qwen3.8-2.4t-a95b.toml @@ -1,12 +1,8 @@ -name = "Qwen3.8 2.4T A95B" -description = "Qwen3.8 2.4T A95B is an open-weight sparse mixture-of-experts model from Qwen and the open-weight variant of Qwen3.8 Max, with 95 billion active parameters out of 2.4 trillion total. Reasoning is always on." -release_date = "2026-08-12" -last_updated = "2026-08-12" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "alibaba/qwen3.8-2.4t-a95b" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 3.4385 @@ -14,10 +10,4 @@ output = 8.59625 cache_read = 0.6877 [limit] -context = 262_144 -input = 262_144 output = 52_429 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.8-27b.toml b/providers/aimlapi/models/alibaba/qwen3.8-27b.toml index 8f9f3b4ff71..77fdf0eda88 100644 --- a/providers/aimlapi/models/alibaba/qwen3.8-27b.toml +++ b/providers/aimlapi/models/alibaba/qwen3.8-27b.toml @@ -1,22 +1,12 @@ -name = "Qwen3.8 27B" -description = "Qwen3.8 27B is an open-weight dense vision-language model from Qwen. It accepts text, image and video inputs and is suited for coding, professional workflows, research, multimodal interaction and long-running agent tasks, with flexible thinking that can be tuned per request." -release_date = "2026-08-14" -last_updated = "2026-08-14" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "alibaba/qwen3.8-27b" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.61893 output = 4.40128 [limit] -context = 262_144 -input = 262_144 output = 131_072 - -[modalities] -input = ["image", "text", "video"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.8-flash.toml b/providers/aimlapi/models/alibaba/qwen3.8-flash.toml index 10fdead5776..811f1426cff 100644 --- a/providers/aimlapi/models/alibaba/qwen3.8-flash.toml +++ b/providers/aimlapi/models/alibaba/qwen3.8-flash.toml @@ -1,23 +1,10 @@ -name = "Qwen3.8 Flash" -description = "Qwen3.8-Flash is the speed- and cost-optimized tier of the Qwen3.8 family, pairing hybrid thinking with high throughput for coding assistance, agentic workflows and visual understanding. Natively multimodal: accepts text, images and video and returns text, with a 1M-token context window. Supports function calling, built-in tools (including web search), structured outputs and context caching." -release_date = "2026-08-26" -last_updated = "2026-08-26" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "alibaba/qwen3.8-flash" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.208 output = 0.611 cache_read = 0.0208 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 131_072 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/alibaba/qwen3.8-max.toml b/providers/aimlapi/models/alibaba/qwen3.8-max.toml index a13befb62e7..654d7093d5e 100644 --- a/providers/aimlapi/models/alibaba/qwen3.8-max.toml +++ b/providers/aimlapi/models/alibaba/qwen3.8-max.toml @@ -1,24 +1,10 @@ -name = "Qwen3.8 Max" -description = "Qwen3.8-Max is Alibaba’s flagship reasoning and agentic LLM, positioned as its highest-capability tier for complex multi-step reasoning, coding and autonomous workflows. Natively multimodal: accepts text and images and returns text, with a 1M-token context window. Supports hybrid thinking mode, function calling, built-in tools (including web search) and structured outputs." -family = "qwen3.8-max" -release_date = "2026-08-03" -last_updated = "2026-08-03" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "alibaba/qwen3.8-max" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 2.6 output = 7.8 cache_read = 0.325 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 131_072 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/anthracite-org/magnum-v4-72b.toml b/providers/aimlapi/models/anthracite-org/magnum-v4-72b.toml deleted file mode 100644 index 515e4cdc003..00000000000 --- a/providers/aimlapi/models/anthracite-org/magnum-v4-72b.toml +++ /dev/null @@ -1,22 +0,0 @@ -name = "Magnum v4 72B" -description = "Magnum V4 is a powerful language model optimized for high-quality text generation tasks." -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 4.1262 -output = 6.877 - -[limit] -context = 32_000 -input = 32_000 -output = 16_384 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-fable-5.toml b/providers/aimlapi/models/anthropic/claude-fable-5.toml deleted file mode 100644 index 7dd55cee867..00000000000 --- a/providers/aimlapi/models/anthropic/claude-fable-5.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Claude Fable 5" -description = "Claude Fable 5 is Anthropic's model designed for complex reasoning and agentic coding workflows. Features a 1M token context window, adaptive thinking, and strong performance across multilingual tasks, long-context handling, and vision." -family = "claude-fable" -release_date = "2026-06-09" -last_updated = "2026-06-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 13 -output = 65 -cache_read = 1.3 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 128_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-haiku-4-5-20251001.toml b/providers/aimlapi/models/anthropic/claude-haiku-4-5-20251001.toml deleted file mode 100644 index ff1d913416f..00000000000 --- a/providers/aimlapi/models/anthropic/claude-haiku-4-5-20251001.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Claude 4.5 Haiku" -description = "Claude 4.5 Haiku is a fast, lightweight model optimized for quick responses and high-throughput applications." -family = "claude-haiku" -release_date = "2025-10-16" -last_updated = "2025-10-16" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.3 -output = 6.5 -cache_read = 0.13 - -[limit] -context = 200_000 -input = 200_000 -output = 64_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-haiku-4.5.toml b/providers/aimlapi/models/anthropic/claude-haiku-4.5.toml index c347be507bf..626bc3d762f 100644 --- a/providers/aimlapi/models/anthropic/claude-haiku-4.5.toml +++ b/providers/aimlapi/models/anthropic/claude-haiku-4.5.toml @@ -1,24 +1,10 @@ -name = "Claude 4.5 Haiku" -description = "Claude 4.5 Haiku is a fast, lightweight model optimized for quick responses and high-throughput applications." -family = "claude-haiku" -release_date = "2025-10-16" -last_updated = "2025-10-16" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "anthropic/claude-haiku-4-5" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 1.3754 output = 6.877 cache_read = 0.13754 - -[limit] -context = 200_000 -input = 200_000 -output = 64_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-opus-4-1-20250805.toml b/providers/aimlapi/models/anthropic/claude-opus-4-1-20250805.toml deleted file mode 100644 index 3da960db23b..00000000000 --- a/providers/aimlapi/models/anthropic/claude-opus-4-1-20250805.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Claude 4.1 Opus" -description = "Claude 4.1 Opus is an enhanced version of Claude 4 Opus with improved reasoning and instruction-following capabilities." -family = "claude-opus" -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 19.5 -output = 97.5 -cache_read = 1.95 - -[limit] -context = 200_000 -input = 200_000 -output = 32_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-opus-4-5-20251101.toml b/providers/aimlapi/models/anthropic/claude-opus-4-5-20251101.toml deleted file mode 100644 index e89d5594097..00000000000 --- a/providers/aimlapi/models/anthropic/claude-opus-4-5-20251101.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Claude 4.5 Opus" -description = "Claude 4.5 Opus is Anthropic's flagship model for the most demanding reasoning and analysis tasks." -family = "claude-opus" -release_date = "2025-11-25" -last_updated = "2025-11-25" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 6.5 -output = 32.5 -cache_read = 0.65 - -[limit] -context = 200_000 -input = 200_000 -output = 64_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-opus-4-6.toml b/providers/aimlapi/models/anthropic/claude-opus-4-6.toml deleted file mode 100644 index fcdb333c25b..00000000000 --- a/providers/aimlapi/models/anthropic/claude-opus-4-6.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Claude 4.6 Opus" -description = "Claude 4.6 Opus delivers Anthropic's highest intelligence with extended output for complex, long-horizon tasks." -family = "claude-opus" -release_date = "2026-02-06" -last_updated = "2026-02-06" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 6.5 -output = 32.5 -cache_read = 0.65 - -[limit] -context = 200_000 -input = 200_000 -output = 128_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-opus-4-8.toml b/providers/aimlapi/models/anthropic/claude-opus-4-8.toml deleted file mode 100644 index 01bf0f0a88f..00000000000 --- a/providers/aimlapi/models/anthropic/claude-opus-4-8.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Claude 4.8 Opus" -description = "Claude 4.8 Opus is Anthropic's most capable model designed for complex reasoning and agentic coding workflows. Features a 1M token context window, adaptive thinking, and top-tier performance across multilingual tasks, long-context handling, and vision." -family = "claude-opus" -release_date = "2026-05-28" -last_updated = "2026-05-28" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 6.5 -output = 32.5 -cache_read = 0.65 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 128_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-opus-4.1.toml b/providers/aimlapi/models/anthropic/claude-opus-4.1.toml index 47786dc4e28..c86491a03be 100644 --- a/providers/aimlapi/models/anthropic/claude-opus-4.1.toml +++ b/providers/aimlapi/models/anthropic/claude-opus-4.1.toml @@ -1,24 +1,10 @@ -name = "Claude 4.1 Opus" -description = "Claude 4.1 Opus is an enhanced version of Claude 4 Opus with improved reasoning and instruction-following capabilities." -family = "claude-opus" -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "anthropic/claude-opus-4-1" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 20.631 output = 103.155 cache_read = 2.0631 - -[limit] -context = 200_000 -input = 200_000 -output = 32_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-opus-4.5.toml b/providers/aimlapi/models/anthropic/claude-opus-4.5.toml index 8a5cb554636..1d5bfafae16 100644 --- a/providers/aimlapi/models/anthropic/claude-opus-4.5.toml +++ b/providers/aimlapi/models/anthropic/claude-opus-4.5.toml @@ -1,24 +1,10 @@ -name = "Claude 4.5 Opus" -description = "Claude 4.5 Opus is Anthropic's flagship model for the most demanding reasoning and analysis tasks." -family = "claude-opus" -release_date = "2025-11-25" -last_updated = "2025-11-25" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "anthropic/claude-opus-4-5" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 6.877 output = 34.385 cache_read = 0.6877 - -[limit] -context = 200_000 -input = 200_000 -output = 64_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-opus-4.7-fast.toml b/providers/aimlapi/models/anthropic/claude-opus-4.7-fast.toml new file mode 100644 index 00000000000..ebcaa448558 --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-opus-4.7-fast.toml @@ -0,0 +1,10 @@ +base_model = "anthropic/claude-opus-4-7" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 41.262 +output = 206.31 +cache_read = 4.1262 diff --git a/providers/aimlapi/models/anthropic/claude-opus-4.7.toml b/providers/aimlapi/models/anthropic/claude-opus-4.7.toml new file mode 100644 index 00000000000..74ee9952772 --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-opus-4.7.toml @@ -0,0 +1,10 @@ +base_model = "anthropic/claude-opus-4-7" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 6.877 +output = 34.385 +cache_read = 0.6877 diff --git a/providers/aimlapi/models/anthropic/claude-opus-4.8-fast.toml b/providers/aimlapi/models/anthropic/claude-opus-4.8-fast.toml new file mode 100644 index 00000000000..eca8426cea0 --- /dev/null +++ b/providers/aimlapi/models/anthropic/claude-opus-4.8-fast.toml @@ -0,0 +1,10 @@ +base_model = "anthropic/claude-opus-4-8" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 13.754 +output = 68.77 +cache_read = 1.3754 diff --git a/providers/aimlapi/models/anthropic/claude-opus-4.8.toml b/providers/aimlapi/models/anthropic/claude-opus-4.8.toml index f724b9df4f8..f57da60fd4f 100644 --- a/providers/aimlapi/models/anthropic/claude-opus-4.8.toml +++ b/providers/aimlapi/models/anthropic/claude-opus-4.8.toml @@ -1,24 +1,10 @@ -name = "Claude 4.8 Opus" -description = "Claude 4.8 Opus is Anthropic's most capable model designed for complex reasoning and agentic coding workflows. Features a 1M token context window, adaptive thinking, and top-tier performance across multilingual tasks, long-context handling, and vision." -family = "claude-opus" -release_date = "2026-05-28" -last_updated = "2026-05-28" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "anthropic/claude-opus-4-8" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 6.877 output = 34.385 cache_read = 0.6877 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-opus-5.toml b/providers/aimlapi/models/anthropic/claude-opus-5.toml deleted file mode 100644 index 5f622b5b070..00000000000 --- a/providers/aimlapi/models/anthropic/claude-opus-5.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Claude Opus 5" -description = "Claude Opus 5 is ideal for complex agentic coding and enterprise work. Features a 1M token context window, adaptive thinking, and top-tier performance across long-horizon agentic tasks, knowledge work, and vision." -family = "claude-opus" -release_date = "2026-07-24" -last_updated = "2026-07-24" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 6.5 -output = 32.5 -cache_read = 0.65 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 128_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-sonnet-4-5-20250929.toml b/providers/aimlapi/models/anthropic/claude-sonnet-4-5-20250929.toml deleted file mode 100644 index 0d392a2e5a5..00000000000 --- a/providers/aimlapi/models/anthropic/claude-sonnet-4-5-20250929.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Claude 4.5 Sonnet" -description = "Claude 4.5 Sonnet offers a balance of intelligence and speed for diverse enterprise and coding tasks." -family = "claude-sonnet" -release_date = "2025-10-01" -last_updated = "2025-10-01" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 3.9 -output = 19.5 -cache_read = 0.39 - -[limit] -context = 200_000 -input = 200_000 -output = 64_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-sonnet-4-6.toml b/providers/aimlapi/models/anthropic/claude-sonnet-4-6.toml deleted file mode 100644 index 4a33fa60413..00000000000 --- a/providers/aimlapi/models/anthropic/claude-sonnet-4-6.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Claude 4.6 Sonnet" -description = "Claude 4.6 Sonnet is a highly capable model balancing performance and efficiency for everyday AI tasks." -family = "claude-sonnet" -release_date = "2026-02-18" -last_updated = "2026-02-18" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 3.9 -output = 19.5 -cache_read = 0.39 - -[limit] -context = 200_000 -input = 200_000 -output = 64_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-sonnet-4.5.toml b/providers/aimlapi/models/anthropic/claude-sonnet-4.5.toml index 4bb067b0250..abdf9b07883 100644 --- a/providers/aimlapi/models/anthropic/claude-sonnet-4.5.toml +++ b/providers/aimlapi/models/anthropic/claude-sonnet-4.5.toml @@ -1,24 +1,10 @@ -name = "Claude 4.5 Sonnet" -description = "Claude 4.5 Sonnet offers a balance of intelligence and speed for diverse enterprise and coding tasks." -family = "claude-sonnet" -release_date = "2025-10-01" -last_updated = "2025-10-01" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "anthropic/claude-sonnet-4-5" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 4.1262 output = 20.631 cache_read = 0.41262 - -[limit] -context = 200_000 -input = 200_000 -output = 64_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-sonnet-4.6.toml b/providers/aimlapi/models/anthropic/claude-sonnet-4.6.toml index 17417c590c9..a46dbba1d12 100644 --- a/providers/aimlapi/models/anthropic/claude-sonnet-4.6.toml +++ b/providers/aimlapi/models/anthropic/claude-sonnet-4.6.toml @@ -1,13 +1,8 @@ -name = "Claude 4.6 Sonnet" -description = "Claude 4.6 Sonnet is a highly capable model balancing performance and efficiency for everyday AI tasks." -family = "claude-sonnet" -release_date = "2026-02-18" -last_updated = "2026-02-18" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "anthropic/claude-sonnet-4-6" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 4.1262 @@ -16,9 +11,3 @@ cache_read = 0.41262 [limit] context = 200_000 -input = 200_000 -output = 64_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/anthropic/claude-sonnet-5.toml b/providers/aimlapi/models/anthropic/claude-sonnet-5.toml deleted file mode 100644 index 07b90a1f06b..00000000000 --- a/providers/aimlapi/models/anthropic/claude-sonnet-5.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Claude Sonnet 5" -description = "Claude Sonnet 5 is a high-performance model for coding and agentic workflows, delivering near-Opus quality at Sonnet cost. Features a 1M token context window, adaptive thinking, and high-resolution vision." -family = "claude-sonnet" -release_date = "2026-06-30" -last_updated = "2026-06-30" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 2.6 -output = 13 -cache_read = 0.26 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 128_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/arcee-ai/trinity-large-thinking.toml b/providers/aimlapi/models/arcee-ai/trinity-large-thinking.toml new file mode 100644 index 00000000000..5d0980ba3eb --- /dev/null +++ b/providers/aimlapi/models/arcee-ai/trinity-large-thinking.toml @@ -0,0 +1,14 @@ +base_model = "arcee-ai/trinity-large-thinking" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.302588 +output = 1.16909 +cache_read = 0.082524 + +[limit] +context = 262_144 +output = 80_000 diff --git a/providers/aimlapi/models/baidu/ernie-4.5-vl-424b-a47b.toml b/providers/aimlapi/models/baidu/ernie-4.5-vl-424b-a47b.toml deleted file mode 100644 index 0fd7d23d03e..00000000000 --- a/providers/aimlapi/models/baidu/ernie-4.5-vl-424b-a47b.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Ernie 4.5 VL 424B A47B" -description = "High-capacity vision-language Ernie 4.5 VL 424B A47B for demanding multimodal workflows." -family = "ernie" -release_date = "2026-01-12" -last_updated = "2026-01-12" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.5577 -output = 1.677 - -[limit] -context = 123_000 -input = 123_000 -output = 123_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/baidu/ernie-5.0.toml b/providers/aimlapi/models/baidu/ernie-5.0.toml deleted file mode 100644 index 0ced30896df..00000000000 --- a/providers/aimlapi/models/baidu/ernie-5.0.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "ERNIE 5.0" -description = "ERNIE 5.0 is Baidu’s unified foundation model supporting text generation, visual understanding, and deep-thinking reasoning, with a 128K context window." -family = "ernie" -release_date = "2026-06-30" -last_updated = "2026-06-30" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.82 -output = 7.28 - -[limit] -context = 128_000 -input = 128_000 -output = 65_536 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/bytedance-seed/seed-2-1-turbo.toml b/providers/aimlapi/models/bytedance-seed/seed-2-1-turbo.toml deleted file mode 100644 index 4b754c1c105..00000000000 --- a/providers/aimlapi/models/bytedance-seed/seed-2-1-turbo.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Seed 2.1 Turbo" -description = "ByteDance Seed 2.1 Turbo is a multimodal model for coding and long-horizon agent workflows, accepting text, image and video input across a 256K context window." -family = "seed" -release_date = "2026-08-10" -last_updated = "2026-08-10" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.6877 -output = 3.4385 - -[limit] -context = 262_144 -input = 262_144 -output = 262_144 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/bytedance-seed/seed-2.0-code.toml b/providers/aimlapi/models/bytedance-seed/seed-2.0-code.toml index 8aeb284b365..b22d70d2be7 100644 --- a/providers/aimlapi/models/bytedance-seed/seed-2.0-code.toml +++ b/providers/aimlapi/models/bytedance-seed/seed-2.0-code.toml @@ -1,23 +1,9 @@ -name = "Seed-2.0-Code" -description = "ByteDance Seed-2.0-Code is optimized for agentic coding — frontend development, multilingual programming tasks and coding-agent workflows — with a 256K context window, multimodal (text/image/video) input and selectable reasoning effort." -family = "seed" -release_date = "2026-07-30" -last_updated = "2026-07-30" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "bytedance-seed/seed-2.0-code" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.6877 output = 4.1262 - -[limit] -context = 262_144 -input = 262_144 -output = 131_072 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/bytedance-seed/seed-2.0-lite.toml b/providers/aimlapi/models/bytedance-seed/seed-2.0-lite.toml new file mode 100644 index 00000000000..06c0e91575d --- /dev/null +++ b/providers/aimlapi/models/bytedance-seed/seed-2.0-lite.toml @@ -0,0 +1,13 @@ +base_model = "bytedance-seed/seed-2.0-lite" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.34385 +output = 2.7508 + +[limit] +context = 262_144 +output = 131_072 diff --git a/providers/aimlapi/models/bytedance-seed/seed-2.0-mini.toml b/providers/aimlapi/models/bytedance-seed/seed-2.0-mini.toml new file mode 100644 index 00000000000..d2cf473b3d6 --- /dev/null +++ b/providers/aimlapi/models/bytedance-seed/seed-2.0-mini.toml @@ -0,0 +1,13 @@ +base_model = "bytedance-seed/seed-2.0-mini" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.13754 +output = 0.55016 + +[limit] +context = 262_144 +output = 131_072 diff --git a/providers/aimlapi/models/bytedance/seed-1-8.toml b/providers/aimlapi/models/bytedance/seed-1-8.toml deleted file mode 100644 index 44b4b7aa6be..00000000000 --- a/providers/aimlapi/models/bytedance/seed-1-8.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Seed 1.8" -description = "ByteDance Seed 1.8 is a large language model with a 256K context window optimized for long-context reasoning and generation." -family = "seed" -release_date = "2026-01-12" -last_updated = "2026-01-12" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.325 -output = 2.6 - -[limit] -context = 256_000 -input = 256_000 -output = 224_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/cohere/command-a.toml b/providers/aimlapi/models/cohere/command-a.toml deleted file mode 100644 index 6733854fea4..00000000000 --- a/providers/aimlapi/models/cohere/command-a.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Command A" -description = "Command A is a powerful LLM with advanced capabilities for enterprise applications." -family = "command-a" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 3.4385 -output = 13.754 - -[limit] -context = 256_000 -input = 256_000 -output = 256_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/cohere/command-r-08-2024.toml b/providers/aimlapi/models/cohere/command-r-08-2024.toml new file mode 100644 index 00000000000..93418a9007b --- /dev/null +++ b/providers/aimlapi/models/cohere/command-r-08-2024.toml @@ -0,0 +1,5 @@ +base_model = "cohere/command-r-08-2024" + +[cost] +input = 0.20631 +output = 0.82524 diff --git a/providers/aimlapi/models/cohere/command-r-plus-08-2024.toml b/providers/aimlapi/models/cohere/command-r-plus-08-2024.toml new file mode 100644 index 00000000000..a7ed695ea2c --- /dev/null +++ b/providers/aimlapi/models/cohere/command-r-plus-08-2024.toml @@ -0,0 +1,5 @@ +base_model = "cohere/command-r-plus-08-2024" + +[cost] +input = 3.4385 +output = 13.754 diff --git a/providers/aimlapi/models/cohere/command-r7b-12-2024.toml b/providers/aimlapi/models/cohere/command-r7b-12-2024.toml new file mode 100644 index 00000000000..6a0ae5ffe27 --- /dev/null +++ b/providers/aimlapi/models/cohere/command-r7b-12-2024.toml @@ -0,0 +1,5 @@ +base_model = "cohere/command-r7b-12-2024" + +[cost] +input = 0.0515775 +output = 0.20631 diff --git a/providers/aimlapi/models/deepseek/deepseek-chat-v3.1.toml b/providers/aimlapi/models/deepseek/deepseek-chat-v3.1.toml deleted file mode 100644 index c2f784804bc..00000000000 --- a/providers/aimlapi/models/deepseek/deepseek-chat-v3.1.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "DeepSeek V3.1" -description = "DeepSeek-V3.1 is the latest advanced LLM with improved architecture and performance across various natural language tasks." -family = "deepseek" -release_date = "2025-08-28" -last_updated = "2025-08-28" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.89401 -output = 2.33818 -cache_read = 0.75647 - -[limit] -context = 128_000 -input = 128_000 -output = 8_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-chat.toml b/providers/aimlapi/models/deepseek/deepseek-chat.toml index 1136d2f3adb..f47b52b0ab1 100644 --- a/providers/aimlapi/models/deepseek/deepseek-chat.toml +++ b/providers/aimlapi/models/deepseek/deepseek-chat.toml @@ -1,13 +1,4 @@ -name = "DeepSeek V3" -description = "DeepSeek-V3 is an advanced LLM with efficient architecture and high performance across various natural language tasks." -family = "deepseek" -release_date = "2025-08-28" -last_updated = "2025-08-28" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "deepseek/deepseek-chat" [cost] input = 0.364 @@ -16,9 +7,4 @@ cache_read = 0.0364 [limit] context = 128_000 -input = 128_000 output = 124_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-non-reasoner-v3.1-terminus.toml b/providers/aimlapi/models/deepseek/deepseek-non-reasoner-v3.1-terminus.toml deleted file mode 100644 index ff1f0cc6afc..00000000000 --- a/providers/aimlapi/models/deepseek/deepseek-non-reasoner-v3.1-terminus.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "DeepSeek V3.1" -description = "DeepSeek-V3.1 is the latest advanced LLM with improved architecture and performance across various natural language tasks." -family = "deepseek" -release_date = "2025-08-28" -last_updated = "2025-08-28" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.371358 -output = 1.3754 -cache_read = 0.371358 - -[limit] -context = 128_000 -input = 128_000 -output = 8_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-non-thinking-v3.2-exp.toml b/providers/aimlapi/models/deepseek/deepseek-non-thinking-v3.2-exp.toml deleted file mode 100644 index e283befe929..00000000000 --- a/providers/aimlapi/models/deepseek/deepseek-non-thinking-v3.2-exp.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "DeepSeek V3.2 Exp" -description = "DeepSeek-V3.2-Exp is an experimental variant of DeepSeek-V3.2 with extended context and non-thinking mode for fast responses." -family = "deepseek" -release_date = "2025-10-03" -last_updated = "2025-10-03" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.371358 -output = 0.563914 -cache_read = 0.371358 - -[limit] -context = 128_000 -input = 128_000 -output = 64_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-r1.toml b/providers/aimlapi/models/deepseek/deepseek-r1.toml deleted file mode 100644 index 0bcdc0d420f..00000000000 --- a/providers/aimlapi/models/deepseek/deepseek-r1.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "DeepSeek Reasoner V3.1" -description = "DeepSeek Reasoner V3.1 excels in reasoning tasks with advanced chain-of-thought processing and the latest improvements in parameter activation." -family = "deepseek" -release_date = "2025-08-28" -last_updated = "2025-08-28" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.6877 -output = 2.95711 -cache_read = 0.6877 - -[limit] -context = 128_000 -input = 128_000 -output = 64_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-reasoner-v3.1-terminus.toml b/providers/aimlapi/models/deepseek/deepseek-reasoner-v3.1-terminus.toml deleted file mode 100644 index ff1f0cc6afc..00000000000 --- a/providers/aimlapi/models/deepseek/deepseek-reasoner-v3.1-terminus.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "DeepSeek V3.1" -description = "DeepSeek-V3.1 is the latest advanced LLM with improved architecture and performance across various natural language tasks." -family = "deepseek" -release_date = "2025-08-28" -last_updated = "2025-08-28" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.371358 -output = 1.3754 -cache_read = 0.371358 - -[limit] -context = 128_000 -input = 128_000 -output = 8_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-reasoner.toml b/providers/aimlapi/models/deepseek/deepseek-reasoner.toml deleted file mode 100644 index eee838b7579..00000000000 --- a/providers/aimlapi/models/deepseek/deepseek-reasoner.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "DeepSeek R1" -description = "DeepSeek R1 excels in reasoning tasks with advanced features like chain-of-thought processing and efficient parameter activation" -family = "deepseek" -release_date = "2025-08-28" -last_updated = "2025-08-28" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.364 -output = 0.546 -cache_read = 0.0364 - -[limit] -context = 128_000 -input = 128_000 -output = 127_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-thinking-v3.2-exp.toml b/providers/aimlapi/models/deepseek/deepseek-thinking-v3.2-exp.toml deleted file mode 100644 index 49d2c75a6c6..00000000000 --- a/providers/aimlapi/models/deepseek/deepseek-thinking-v3.2-exp.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "DeepSeek V3.2 Exp" -description = "DeepSeek-V3.2-Exp is an experimental variant of DeepSeek-V3.2 with extended context and non-thinking mode for fast responses." -family = "deepseek-thinking" -release_date = "2025-10-03" -last_updated = "2025-10-03" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.371358 -output = 0.563914 -cache_read = 0.371358 - -[limit] -context = 128_000 -input = 128_000 -output = 64_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-v4-flash-vision-exp.toml b/providers/aimlapi/models/deepseek/deepseek-v4-flash-vision-exp.toml index e14ec62d5c9..14c8b06da73 100644 --- a/providers/aimlapi/models/deepseek/deepseek-v4-flash-vision-exp.toml +++ b/providers/aimlapi/models/deepseek/deepseek-v4-flash-vision-exp.toml @@ -1,13 +1,8 @@ -name = "DeepSeek V4 Flash Vision Exp" -description = "DeepSeek V4 Flash Vision Exp is an experimental vision-enabled build of DeepSeek V4 Flash. It adds image understanding on top of the base model while matching it on text tasks, with a 1M-token context window and thinking mode enabled by default." -family = "deepseek" -release_date = "2026-08-21" -last_updated = "2026-08-21" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "deepseek/deepseek-v4-flash-vision-exp" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.572 @@ -16,9 +11,3 @@ cache_read = 0.0182 [limit] context = 1_048_576 -input = 1_048_576 -output = 384_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-v4-flash.toml b/providers/aimlapi/models/deepseek/deepseek-v4-flash.toml index 822af6bc3cd..56437429724 100644 --- a/providers/aimlapi/models/deepseek/deepseek-v4-flash.toml +++ b/providers/aimlapi/models/deepseek/deepseek-v4-flash.toml @@ -1,24 +1,10 @@ -name = "DeepSeek V4 Flash" -description = "DeepSeek V4 Flash and Pro are large language models for chat and reasoning tasks with up to 1M context length. They support both standard and thinking modes, enabling efficient responses or deeper reasoning workflows." -family = "deepseek" -release_date = "2026-04-24" -last_updated = "2026-04-24" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "deepseek/deepseek-v4-flash" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.182 output = 0.364 cache_read = 0.00364 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 384_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-v4-pro-0813.toml b/providers/aimlapi/models/deepseek/deepseek-v4-pro-0813.toml index 710bd150065..c8947cb45c0 100644 --- a/providers/aimlapi/models/deepseek/deepseek-v4-pro-0813.toml +++ b/providers/aimlapi/models/deepseek/deepseek-v4-pro-0813.toml @@ -1,13 +1,8 @@ -name = "DeepSeek V4 Pro 0813" -description = "DeepSeek V4 Pro 0813 is the GA release of DeepSeek V4 Pro — a large-scale mixture-of-experts model for chat and reasoning tasks with 1M context length, supporting both standard and thinking modes." -family = "deepseek" -release_date = "2026-08-12" -last_updated = "2026-08-12" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "deepseek/deepseek-v4-pro-0813" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 2.393196 @@ -16,9 +11,3 @@ cache_read = 0.199433 [limit] context = 1_048_576 -input = 1_048_576 -output = 384_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/deepseek/deepseek-v4-pro.toml b/providers/aimlapi/models/deepseek/deepseek-v4-pro.toml index 0494f280ab4..e733d802ed8 100644 --- a/providers/aimlapi/models/deepseek/deepseek-v4-pro.toml +++ b/providers/aimlapi/models/deepseek/deepseek-v4-pro.toml @@ -1,24 +1,10 @@ -name = "DeepSeek V4 Pro" -description = "DeepSeek V4 Flash and Pro are large language models for chat and reasoning tasks with up to 1M context length. They support both standard and thinking modes, enabling efficient responses or deeper reasoning workflows." -family = "deepseek" -release_date = "2026-04-24" -last_updated = "2026-04-24" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "deepseek/deepseek-v4-pro" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.5655 output = 1.131 cache_read = 0.0047125 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 384_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-2.5-flash-image.toml b/providers/aimlapi/models/google/gemini-2.5-flash-image.toml new file mode 100644 index 00000000000..cd9fc31bd11 --- /dev/null +++ b/providers/aimlapi/models/google/gemini-2.5-flash-image.toml @@ -0,0 +1,10 @@ +base_model = "google/gemini-2.5-flash-image" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.742716 +output = 6.1893 +cache_read = 0.0742716 diff --git a/providers/aimlapi/models/google/gemini-2.5-flash-lite-preview.toml b/providers/aimlapi/models/google/gemini-2.5-flash-lite-preview.toml deleted file mode 100644 index 015b166c9aa..00000000000 --- a/providers/aimlapi/models/google/gemini-2.5-flash-lite-preview.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Gemini 2.5 Flash Lite Preview" -description = "Gemini 2.5 Flash Lite Preview is a lightweight AI model developed by Google, optimized for quick responses and efficient processing, making it ideal for tasks requiring minimal latency and resource consumption" -family = "gemini" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.0975 -output = 0.39 -cache_read = 0.0975 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 1_048_576 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-2.5-flash-lite.toml b/providers/aimlapi/models/google/gemini-2.5-flash-lite.toml index 015b166c9aa..84c6d6a63de 100644 --- a/providers/aimlapi/models/google/gemini-2.5-flash-lite.toml +++ b/providers/aimlapi/models/google/gemini-2.5-flash-lite.toml @@ -1,13 +1,8 @@ -name = "Gemini 2.5 Flash Lite Preview" -description = "Gemini 2.5 Flash Lite Preview is a lightweight AI model developed by Google, optimized for quick responses and efficient processing, making it ideal for tasks requiring minimal latency and resource consumption" -family = "gemini" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "google/gemini-2.5-flash-lite" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.0975 @@ -16,9 +11,4 @@ cache_read = 0.0975 [limit] context = 1_000_000 -input = 1_000_000 output = 1_048_576 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-2.5-pro.toml b/providers/aimlapi/models/google/gemini-2.5-pro.toml deleted file mode 100644 index 7729dd62370..00000000000 --- a/providers/aimlapi/models/google/gemini-2.5-pro.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Gemini 2.5 Pro" -description = "Gemini Pro 2.5 is Google's most advanced reasoning AI model designed for complex tasks, showcasing strong reasoning and coding capabilities. It leads various benchmarks, indicating its enhanced performance." -family = "gemini" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.625 -output = 13 -cache_read = 1.625 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 65_536 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-3-6-flash.toml b/providers/aimlapi/models/google/gemini-3-6-flash.toml deleted file mode 100644 index 8326f7dfdf9..00000000000 --- a/providers/aimlapi/models/google/gemini-3-6-flash.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Gemini 3.6 Flash" -description = "Gemini 3.6 Flash is Google’s most intelligent Flash model, balancing speed with frontier intelligence for strong performance on agentic, coding and multimodal tasks, with superior search and grounding." -family = "gemini" -release_date = "2026-07-21" -last_updated = "2026-07-21" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.95 -output = 9.75 -cache_read = 0.195 - -[limit] -context = 1_048_576 -input = 1_048_576 -output = 65_536 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-3-7-flash.toml b/providers/aimlapi/models/google/gemini-3-7-flash.toml deleted file mode 100644 index d8dbbdc7314..00000000000 --- a/providers/aimlapi/models/google/gemini-3-7-flash.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Gemini 3.7 Flash" -description = "Gemini 3.7 Flash is the high-efficiency Flash model of the Gemini 3 family, with Pro-level agentic capabilities, stronger code generation and terminal execution, and high token efficiency for multi-step multimodal work." -family = "gemini" -release_date = "2026-08-13" -last_updated = "2026-08-13" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.975 -output = 4.875 -cache_read = 0.0975 - -[limit] -context = 1_048_576 -input = 1_048_576 -output = 65_536 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-3-flash-preview.toml b/providers/aimlapi/models/google/gemini-3-flash-preview.toml deleted file mode 100644 index 1005f2ee312..00000000000 --- a/providers/aimlapi/models/google/gemini-3-flash-preview.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Gemini 3 Flash Preview" -description = "Gemini 3 Flash Preview is a fast, efficient variant of Gemini 3 designed for low-latency multimodal applications." -family = "gemini" -release_date = "2025-12-17" -last_updated = "2025-12-17" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.65 -output = 3.9 -cache_read = 0.65 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 65_536 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-3-pro-image-preview.toml b/providers/aimlapi/models/google/gemini-3-pro-image-preview.toml new file mode 100644 index 00000000000..4932eb4ba0e --- /dev/null +++ b/providers/aimlapi/models/google/gemini-3-pro-image-preview.toml @@ -0,0 +1,10 @@ +base_model = "google/gemini-3-pro-image-preview" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 2.7508 +output = 16.5048 +cache_read = 0.27508 diff --git a/providers/aimlapi/models/google/gemini-3-pro-image.toml b/providers/aimlapi/models/google/gemini-3-pro-image.toml new file mode 100644 index 00000000000..7f6c33369c9 --- /dev/null +++ b/providers/aimlapi/models/google/gemini-3-pro-image.toml @@ -0,0 +1,10 @@ +base_model = "google/gemini-3-pro-image" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 2.7508 +output = 16.5048 +cache_read = 0.27508 diff --git a/providers/aimlapi/models/google/gemini-3.1-flash-image-preview.toml b/providers/aimlapi/models/google/gemini-3.1-flash-image-preview.toml new file mode 100644 index 00000000000..19d5157aacd --- /dev/null +++ b/providers/aimlapi/models/google/gemini-3.1-flash-image-preview.toml @@ -0,0 +1,13 @@ +base_model = "google/gemini-3.1-flash-image-preview" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.6877 +output = 4.1262 + +[limit] +context = 131_072 +output = 32_768 diff --git a/providers/aimlapi/models/google/gemini-3.1-flash-image.toml b/providers/aimlapi/models/google/gemini-3.1-flash-image.toml new file mode 100644 index 00000000000..6856e087191 --- /dev/null +++ b/providers/aimlapi/models/google/gemini-3.1-flash-image.toml @@ -0,0 +1,9 @@ +base_model = "google/gemini-3.1-flash-image" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.6877 +output = 4.1262 diff --git a/providers/aimlapi/models/google/gemini-3.1-flash-lite-image.toml b/providers/aimlapi/models/google/gemini-3.1-flash-lite-image.toml new file mode 100644 index 00000000000..c9a308a289e --- /dev/null +++ b/providers/aimlapi/models/google/gemini-3.1-flash-lite-image.toml @@ -0,0 +1,12 @@ +base_model = "google/gemini-3.1-flash-lite-image" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.34385 +output = 2.0631 + +[limit] +output = 66_000 diff --git a/providers/aimlapi/models/google/gemini-3.1-flash-lite-preview.toml b/providers/aimlapi/models/google/gemini-3.1-flash-lite-preview.toml new file mode 100644 index 00000000000..e4507cc7a27 --- /dev/null +++ b/providers/aimlapi/models/google/gemini-3.1-flash-lite-preview.toml @@ -0,0 +1,13 @@ +base_model = "google/gemini-3.1-flash-lite-preview" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.34385 +output = 2.0631 +cache_read = 0.034385 + +[limit] +context = 1_000_000 diff --git a/providers/aimlapi/models/google/gemini-3.1-pro-preview-customtools.toml b/providers/aimlapi/models/google/gemini-3.1-pro-preview-customtools.toml new file mode 100644 index 00000000000..090f8c1040b --- /dev/null +++ b/providers/aimlapi/models/google/gemini-3.1-pro-preview-customtools.toml @@ -0,0 +1,13 @@ +base_model = "google/gemini-3.1-pro-preview-customtools" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 2.7508 +output = 16.5048 +cache_read = 0.27508 + +[limit] +context = 1_048_756 diff --git a/providers/aimlapi/models/google/gemini-3.1-pro-preview.toml b/providers/aimlapi/models/google/gemini-3.1-pro-preview.toml deleted file mode 100644 index f79176aabe2..00000000000 --- a/providers/aimlapi/models/google/gemini-3.1-pro-preview.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Gemini 3.1 Pro Preview" -description = "A frontier reasoning model optimized for software engineering and agentic workflows with 1M token context." -family = "gemini" -release_date = "2026-02-25" -last_updated = "2026-02-25" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 2.6 -output = 15.6 -cache_read = 0.65 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 65_536 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-3.6-flash.toml b/providers/aimlapi/models/google/gemini-3.6-flash.toml index 8326f7dfdf9..8cce4b4e5db 100644 --- a/providers/aimlapi/models/google/gemini-3.6-flash.toml +++ b/providers/aimlapi/models/google/gemini-3.6-flash.toml @@ -1,24 +1,10 @@ -name = "Gemini 3.6 Flash" -description = "Gemini 3.6 Flash is Google’s most intelligent Flash model, balancing speed with frontier intelligence for strong performance on agentic, coding and multimodal tasks, with superior search and grounding." -family = "gemini" -release_date = "2026-07-21" -last_updated = "2026-07-21" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "google/gemini-3.6-flash" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 1.95 output = 9.75 cache_read = 0.195 - -[limit] -context = 1_048_576 -input = 1_048_576 -output = 65_536 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-3.7-flash.toml b/providers/aimlapi/models/google/gemini-3.7-flash.toml index d8dbbdc7314..c159244eee9 100644 --- a/providers/aimlapi/models/google/gemini-3.7-flash.toml +++ b/providers/aimlapi/models/google/gemini-3.7-flash.toml @@ -1,24 +1,10 @@ -name = "Gemini 3.7 Flash" -description = "Gemini 3.7 Flash is the high-efficiency Flash model of the Gemini 3 family, with Pro-level agentic capabilities, stronger code generation and terminal execution, and high token efficiency for multi-step multimodal work." -family = "gemini" -release_date = "2026-08-13" -last_updated = "2026-08-13" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "google/gemini-3.7-flash" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.975 output = 4.875 cache_read = 0.0975 - -[limit] -context = 1_048_576 -input = 1_048_576 -output = 65_536 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/google/gemini-flash-latest.toml b/providers/aimlapi/models/google/gemini-flash-latest.toml new file mode 100644 index 00000000000..65342ef400e --- /dev/null +++ b/providers/aimlapi/models/google/gemini-flash-latest.toml @@ -0,0 +1,10 @@ +base_model = "google/gemini-flash-latest" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 2.0631 +output = 12.3786 +cache_read = 0.20631 diff --git a/providers/aimlapi/models/google/gemma-3-12b-it.toml b/providers/aimlapi/models/google/gemma-3-12b-it.toml deleted file mode 100644 index 220d523bfaf..00000000000 --- a/providers/aimlapi/models/google/gemma-3-12b-it.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Gemma 3 12B" -description = "Gemma 3 12B is a mid-size open language model from Google offering a balance of capability and efficiency." -family = "gemma" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.06877 -output = 0.20631 -cache_read = 0.055016 - -[limit] -context = 131_000 -input = 131_000 -output = 96_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/google/gemma-3-27b-it.toml b/providers/aimlapi/models/google/gemma-3-27b-it.toml deleted file mode 100644 index 14469a91d3e..00000000000 --- a/providers/aimlapi/models/google/gemma-3-27b-it.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Gemma 3 27B" -description = "Gemma 3 27B is Google's largest Gemma open model, delivering strong reasoning and text generation capabilities." -family = "gemma" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.20631 -output = 0.632684 -cache_read = 0.103155 - -[limit] -context = 128_000 -input = 128_000 -output = 131_072 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/google/gemma-3-4b-it.toml b/providers/aimlapi/models/google/gemma-3-4b-it.toml deleted file mode 100644 index fae7ae3c2ff..00000000000 --- a/providers/aimlapi/models/google/gemma-3-4b-it.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Gemma 3 4B" -description = "Gemma 3 4B is a lightweight open language model from Google, suitable for on-device deployment and efficient inference." -family = "gemma" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.06877 -output = 0.13754 -cache_read = 0.055016 - -[limit] -context = 131_000 -input = 131_000 -output = 32_768 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/google/gemma-4-26b-a4b-it-maas.toml b/providers/aimlapi/models/google/gemma-4-26b-a4b-it-maas.toml deleted file mode 100644 index 276126dd742..00000000000 --- a/providers/aimlapi/models/google/gemma-4-26b-a4b-it-maas.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Gemma 4 26B A4B IT" -description = "Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts open model from Google." -family = "gemma" -release_date = "2026-06-23" -last_updated = "2026-06-23" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.195 -output = 0.78 -cache_read = 0.195 - -[limit] -context = 262_144 -input = 262_144 -output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/google/gemma-4-26b-a4b-it.toml b/providers/aimlapi/models/google/gemma-4-26b-a4b-it.toml index 953a8582058..86370326cb4 100644 --- a/providers/aimlapi/models/google/gemma-4-26b-a4b-it.toml +++ b/providers/aimlapi/models/google/gemma-4-26b-a4b-it.toml @@ -1,13 +1,8 @@ -name = "Gemma 4 26B A4B IT" -description = "Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts open model from Google." -family = "gemma" -release_date = "2026-06-23" -last_updated = "2026-06-23" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "google/gemma-4-26b-a4b-it" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.20631 @@ -15,10 +10,4 @@ output = 0.82524 cache_read = 0.20631 [limit] -context = 262_144 -input = 262_144 output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/google/gemma-4-31b-it.toml b/providers/aimlapi/models/google/gemma-4-31b-it.toml new file mode 100644 index 00000000000..19755a0cfa3 --- /dev/null +++ b/providers/aimlapi/models/google/gemma-4-31b-it.toml @@ -0,0 +1,14 @@ +base_model = "google/gemma-4-31b-it" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 1.361646 +output = 2.049346 +cache_read = 1.361646 + +[limit] +context = 256_000 +output = 262_144 diff --git a/providers/aimlapi/models/gryphe/mythomax-l2-13b.toml b/providers/aimlapi/models/gryphe/mythomax-l2-13b.toml deleted file mode 100644 index a713feabe4d..00000000000 --- a/providers/aimlapi/models/gryphe/mythomax-l2-13b.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "MythoMax 13B" -description = "One of the highest performing and most popular fine-tunes of Llama 2 13B, with rich descriptions and roleplay." -release_date = "2025-12-09" -last_updated = "2025-12-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 2.578875 -output = 2.578875 -cache_read = 2.578875 - -[limit] -context = 4_096 -input = 4_096 -output = 4_096 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/meituan/longcat-2.0.toml b/providers/aimlapi/models/meituan/longcat-2.0.toml deleted file mode 100644 index fae4399e81d..00000000000 --- a/providers/aimlapi/models/meituan/longcat-2.0.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "LongCat-2.0" -description = "LongCat-2.0 is Meituan's trillion-parameter agentic MoE model with a 1M-token context window and up to 128K output tokens. It supports native tool calling and multi-step reasoning." -family = "longcat" -release_date = "2026-06-30" -last_updated = "2026-06-30" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.975 -output = 3.8349999999999995 -cache_read = 0.0195 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 128_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/meta-llama/Llama-3.3-70B-Instruct-Turbo.toml b/providers/aimlapi/models/meta-llama/Llama-3.3-70B-Instruct-Turbo.toml deleted file mode 100644 index 6c8422c2c86..00000000000 --- a/providers/aimlapi/models/meta-llama/Llama-3.3-70B-Instruct-Turbo.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Llama 3.3 70B Instruct Turbo" -description = "Meta Llama 3.3 70B Instruct Turbo is an advanced language model optimized for instruction-following tasks with high efficiency and performance." -family = "llama" -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.144 -output = 1.144 - -[limit] -context = 128_000 -input = 128_000 -output = 127_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/meta-llama/llama-3.3-70b-versatile.toml b/providers/aimlapi/models/meta-llama/llama-3.3-70b-versatile.toml deleted file mode 100644 index 179f2db61c9..00000000000 --- a/providers/aimlapi/models/meta-llama/llama-3.3-70b-versatile.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Llama 3.3 70B Versatile" -description = "Llama-3.3-70B-Versatile is Meta's advanced multilingual large language model, optimized for a wide range of natural language processing tasks. With 70 billion parameters, it offers high performance across various benchmarks while maintaining efficiency suitable for diverse applications." -family = "llama" -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.767 -output = 1.027 - -[limit] -context = 131_072 -input = 131_072 -output = 32_768 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/meta/muse-glimmer-30b.toml b/providers/aimlapi/models/meta/muse-glimmer-30b.toml new file mode 100644 index 00000000000..adfbd696b94 --- /dev/null +++ b/providers/aimlapi/models/meta/muse-glimmer-30b.toml @@ -0,0 +1,10 @@ +base_model = "meta/muse-glimmer-30b" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.48139 +output = 2.0631 +cache_read = 0.055016 diff --git a/providers/aimlapi/models/meta/muse-spark-1.1.toml b/providers/aimlapi/models/meta/muse-spark-1.1.toml new file mode 100644 index 00000000000..1cf23afdac9 --- /dev/null +++ b/providers/aimlapi/models/meta/muse-spark-1.1.toml @@ -0,0 +1,13 @@ +base_model = "meta/muse-spark-1.1" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 1.71925 +output = 5.84545 +cache_read = 0.20631 + +[limit] +context = 1_048_576 diff --git a/providers/aimlapi/models/meta/muse-spark-1.2.toml b/providers/aimlapi/models/meta/muse-spark-1.2.toml new file mode 100644 index 00000000000..089d86b4862 --- /dev/null +++ b/providers/aimlapi/models/meta/muse-spark-1.2.toml @@ -0,0 +1,10 @@ +base_model = "meta/muse-spark-1.2" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 1.71925 +output = 5.84545 +cache_read = 0.20631 diff --git a/providers/aimlapi/models/minimax/MiniMax-Text-01.toml b/providers/aimlapi/models/minimax/MiniMax-Text-01.toml deleted file mode 100644 index 132810f4fc0..00000000000 --- a/providers/aimlapi/models/minimax/MiniMax-Text-01.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Text 01" -description = "MiniMax-Text-01 excels in long-context processing with advanced features like hybrid attention mechanisms and open-source accessibility." -family = "minimax" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.26 -output = 1.43 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 40_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/minimax/m1.toml b/providers/aimlapi/models/minimax/m1.toml deleted file mode 100644 index d95f3f396eb..00000000000 --- a/providers/aimlapi/models/minimax/m1.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "MiniMax M1" -description = "MiniMax-M1 is a powerful reasoning model with a 1M token context window, delivering strong performance on complex tasks." -family = "minimax" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.52 -output = 2.86 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 40_960 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/minimax/m2-1.toml b/providers/aimlapi/models/minimax/m2-1.toml deleted file mode 100644 index f1a9cff43da..00000000000 --- a/providers/aimlapi/models/minimax/m2-1.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "MiniMax M2.1" -description = "MiniMax-M2.1 is an updated Minimax model with a 200K context window optimized for conversational and reasoning tasks." -family = "minimax" -release_date = "2025-12-23" -last_updated = "2025-12-23" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.39 -output = 1.56 -cache_read = 0.39 - -[limit] -context = 204_000 -input = 204_000 -output = 204_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/minimax/m2-5-20260218.toml b/providers/aimlapi/models/minimax/m2-5-20260218.toml deleted file mode 100644 index 818cb58d735..00000000000 --- a/providers/aimlapi/models/minimax/m2-5-20260218.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "MiniMax M2.5" -description = "MiniMax-M2.5 is a general-purpose large language model by MiniMax designed for text generation and conversational AI use cases." -family = "minimax" -release_date = "2026-02-19" -last_updated = "2026-02-19" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.39 -output = 1.56 -cache_read = 0.39 - -[limit] -context = 204_800 -input = 204_800 -output = 204_800 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/minimax/m2-5-highspeed-20260218.toml b/providers/aimlapi/models/minimax/m2-5-highspeed-20260218.toml deleted file mode 100644 index 6797e39df8a..00000000000 --- a/providers/aimlapi/models/minimax/m2-5-highspeed-20260218.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "MiniMax M2.5 Highspeed" -description = "MiniMax-M2.5 Highspeed is a low-latency, high-throughput version of MiniMax-M2.5 optimized for real-time applications." -family = "minimax" -release_date = "2026-02-19" -last_updated = "2026-02-19" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.78 -output = 3.12 -cache_read = 0.78 - -[limit] -context = 204_800 -input = 204_800 -output = 204_800 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/minimax/m2.toml b/providers/aimlapi/models/minimax/m2.toml deleted file mode 100644 index ad94739969e..00000000000 --- a/providers/aimlapi/models/minimax/m2.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "MiniMax M2" -description = "MiniMax-M2 is an advanced large language model with 1M token context for extended reasoning and generation." -family = "minimax" -release_date = "2025-11-04" -last_updated = "2025-11-04" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.39 -output = 1.56 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 40_960 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/minimax/minimax-m3.toml b/providers/aimlapi/models/minimax/minimax-m3.toml deleted file mode 100644 index be258352180..00000000000 --- a/providers/aimlapi/models/minimax/minimax-m3.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "MiniMax M3" -description = "MiniMax M3 is a large-context reasoning model from MiniMax. Supports up to 1M context tokens and is optimized for long-document understanding, coding workflows, agent tasks, and complex multi-step reasoning." -family = "minimax-m3" -release_date = "2026-06-01" -last_updated = "2026-06-01" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.39 -output = 1.56 -cache_read = 0.078 - -[limit] -context = 524_288 -input = 524_288 -output = 524_288 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/mistralai/devstral-2512.toml b/providers/aimlapi/models/mistralai/devstral-2512.toml new file mode 100644 index 00000000000..1d2bca286c3 --- /dev/null +++ b/providers/aimlapi/models/mistralai/devstral-2512.toml @@ -0,0 +1,6 @@ +base_model = "mistral/devstral-2512" + +[cost] +input = 0.605176 +output = 3.02588 +cache_read = 0.06051759999999999 diff --git a/providers/aimlapi/models/mistralai/mistral-large-2512.toml b/providers/aimlapi/models/mistralai/mistral-large-2512.toml new file mode 100644 index 00000000000..df98e427764 --- /dev/null +++ b/providers/aimlapi/models/mistralai/mistral-large-2512.toml @@ -0,0 +1,6 @@ +base_model = "mistral/mistral-large-2512" + +[cost] +input = 0.6877 +output = 2.0631 +cache_read = 0.06877 diff --git a/providers/aimlapi/models/mistralai/mistral-nemo.toml b/providers/aimlapi/models/mistralai/mistral-nemo.toml index 8a5e3238993..f762b3836f3 100644 --- a/providers/aimlapi/models/mistralai/mistral-nemo.toml +++ b/providers/aimlapi/models/mistralai/mistral-nemo.toml @@ -1,13 +1,4 @@ -name = "Mistral NeMo" -description = "Mistral-Nemo is a powerful multilingual language model with advanced capabilities." -family = "mistral-nemo" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "mistral/mistral-nemo" [cost] input = 0.20631 @@ -15,10 +6,4 @@ output = 0.233818 cache_read = 0.20631 [limit] -context = 128_000 -input = 128_000 output = 131_072 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/mistralai/mistral-small-2603.toml b/providers/aimlapi/models/mistralai/mistral-small-2603.toml new file mode 100644 index 00000000000..fdba8062af4 --- /dev/null +++ b/providers/aimlapi/models/mistralai/mistral-small-2603.toml @@ -0,0 +1,13 @@ +base_model = "mistral/mistral-small-2603" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.2578875 +output = 1.03155 +cache_read = 0.020631 + +[limit] +context = 262_144 diff --git a/providers/aimlapi/models/moonshot/kimi-k2-5.toml b/providers/aimlapi/models/moonshot/kimi-k2-5.toml deleted file mode 100644 index 06b66633c34..00000000000 --- a/providers/aimlapi/models/moonshot/kimi-k2-5.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Kimi K2.5" -description = "Kimi k2.5 is a powerful agentic model by Moonshot AI with strong coding and tool-use capabilities." -family = "kimi-k2" -release_date = "2026-02-06" -last_updated = "2026-02-06" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.78 -output = 3.9 -cache_read = 0.13 - -[limit] -context = 262_144 -input = 262_144 -output = 131_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/moonshot/kimi-k2-7-code-highspeed.toml b/providers/aimlapi/models/moonshot/kimi-k2-7-code-highspeed.toml deleted file mode 100644 index 4387f1e0736..00000000000 --- a/providers/aimlapi/models/moonshot/kimi-k2-7-code-highspeed.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Kimi K2.7 Code Highspeed" -description = "High-throughput variant of Kimi K2.7 Code from Moonshot — the same agentic coding model with extended reasoning and tool use, served at higher speed." -family = "kimi-k2" -release_date = "2026-06-22" -last_updated = "2026-06-22" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 2.47 -output = 10.4 -cache_read = 0.49399999999999994 - -[limit] -context = 262_144 -input = 262_144 -output = 260_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/moonshot/kimi-k2-7-code.toml b/providers/aimlapi/models/moonshot/kimi-k2-7-code.toml deleted file mode 100644 index 11d913e2058..00000000000 --- a/providers/aimlapi/models/moonshot/kimi-k2-7-code.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Kimi K2.7 Code" -description = "Agentic coding model from Moonshot with extended reasoning and tool use." -family = "kimi-k2" -release_date = "2026-06-12" -last_updated = "2026-06-12" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.235 -output = 5.2 -cache_read = 0.24699999999999997 - -[limit] -context = 262_144 -input = 262_144 -output = 260_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/moonshotai/kimi-k2-thinking.toml b/providers/aimlapi/models/moonshotai/kimi-k2-thinking.toml new file mode 100644 index 00000000000..ff971baa25e --- /dev/null +++ b/providers/aimlapi/models/moonshotai/kimi-k2-thinking.toml @@ -0,0 +1,13 @@ +base_model = "moonshotai/kimi-k2-thinking" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.82524 +output = 3.4385 +cache_read = 0.20631 + +[limit] +output = 100_352 diff --git a/providers/aimlapi/models/nousresearch/hermes-4-405b.toml b/providers/aimlapi/models/nousresearch/hermes-4-405b.toml deleted file mode 100644 index 495fa1936df..00000000000 --- a/providers/aimlapi/models/nousresearch/hermes-4-405b.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Hermes 4 405B" -description = "Hermes 4 is a large-scale reasoning model built on Meta-Llama-3.1-405B with hybrid reasoning mode. It can choose to deliberate internally with ... traces or respond directly, offering flexibility between speed and depth. Supports structured outputs, JSON mode, function calling, and tool use." -family = "nousresearch" -release_date = "2025-10-02" -last_updated = "2025-10-02" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.3754 -output = 4.1262 -cache_read = 1.3754 - -[limit] -context = 131_072 -input = 131_072 -output = 16_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/nvidia/nemotron-3-nano-30b-a3b.toml b/providers/aimlapi/models/nvidia/nemotron-3-nano-30b-a3b.toml deleted file mode 100644 index 6f709bd307a..00000000000 --- a/providers/aimlapi/models/nvidia/nemotron-3-nano-30b-a3b.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Nemotron 3 Nano 30B A3B" -description = "NVIDIA Nemotron 3 Nano 30B A3B is a lightweight MoE language model from the Nemotron 3 family, designed for efficient and scalable text generation and agent workflows with strong compute efficiency." -family = "nemotron" -release_date = "2026-04-30" -last_updated = "2026-04-30" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.06877 -output = 0.27508 - -[limit] -context = 262_144 -input = 262_144 -output = 228_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/nvidia/nemotron-3-super-120b-a12b.toml b/providers/aimlapi/models/nvidia/nemotron-3-super-120b-a12b.toml deleted file mode 100644 index 896ed2fc7b5..00000000000 --- a/providers/aimlapi/models/nvidia/nemotron-3-super-120b-a12b.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Nemotron 3 Super 120B A12B" -description = "NVIDIA Nemotron 3 Super 120B A12B is the high-capacity MoE variant of the Nemotron 3 family, delivering strong reasoning and long-context performance for demanding text generation and agent workflows while preserving MoE compute efficiency." -family = "nemotron" -release_date = "2026-04-30" -last_updated = "2026-04-30" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.123786 -output = 0.61893 - -[limit] -context = 262_144 -input = 262_144 -output = 262_144 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/nvidia/nemotron-3-ultra-550b-a55b.toml b/providers/aimlapi/models/nvidia/nemotron-3-ultra-550b-a55b.toml deleted file mode 100644 index cb7b95b096a..00000000000 --- a/providers/aimlapi/models/nvidia/nemotron-3-ultra-550b-a55b.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Nemotron 3 Ultra 550B A55B" -description = "NVIDIA Nemotron 3 Ultra is a reasoning and orchestration model built on a hybrid Transformer-Mamba Mixture-of-Experts architecture, optimized for complex reasoning, long-context analysis, and agent workflows with up to 1M context length." -family = "nemotron" -release_date = "2026-06-04" -last_updated = "2026-06-04" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.82524 -output = 4.95144 -cache_read = 0.20631 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 32_768 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/nvidia/nemotron-3.5-lightning.toml b/providers/aimlapi/models/nvidia/nemotron-3.5-lightning.toml deleted file mode 100644 index 716b6fc15b0..00000000000 --- a/providers/aimlapi/models/nvidia/nemotron-3.5-lightning.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Nemotron 3.5 Lightning" -description = "NVIDIA Nemotron 3.5 Lightning is an open mixture-of-experts model with 3B active parameters out of 30B total, built for high-throughput agentic workloads and specialized task execution with up to 1M context length." -family = "nemotron" -release_date = "2026-08-11" -last_updated = "2026-08-11" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0 -output = 0 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 65_536 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-3.5-turbo.toml b/providers/aimlapi/models/openai/gpt-3.5-turbo.toml new file mode 100644 index 00000000000..aceb5cf5383 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-3.5-turbo.toml @@ -0,0 +1,8 @@ +base_model = "openai/gpt-3.5-turbo" + +[cost] +input = 0.65 +output = 1.95 + +[limit] +context = 16_000 diff --git a/providers/aimlapi/models/openai/gpt-4-turbo.toml b/providers/aimlapi/models/openai/gpt-4-turbo.toml new file mode 100644 index 00000000000..d90a9e47eb8 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-4-turbo.toml @@ -0,0 +1,5 @@ +base_model = "openai/gpt-4-turbo" + +[cost] +input = 13 +output = 39 diff --git a/providers/aimlapi/models/openai/gpt-4.1-2025-04-14.toml b/providers/aimlapi/models/openai/gpt-4.1-2025-04-14.toml deleted file mode 100644 index 75228bcff3b..00000000000 --- a/providers/aimlapi/models/openai/gpt-4.1-2025-04-14.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GPT-4.1" -description = "GPT-4.1 is OpenAI's large-context language model with up to 1M token input and improved instruction following." -family = "gpt" -release_date = "2025-08-07" -last_updated = "2025-08-07" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 2.6 -output = 10.4 -cache_read = 0.65 - -[limit] -context = 1_047_576 -input = 1_047_576 -output = 32_768 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-4.1-mini.toml b/providers/aimlapi/models/openai/gpt-4.1-mini.toml new file mode 100644 index 00000000000..aabb6d5ed5b --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-4.1-mini.toml @@ -0,0 +1,9 @@ +base_model = "openai/gpt-4.1-mini" + +[cost] +input = 0.52 +output = 2.08 +cache_read = 0.13 + +[limit] +context = 1_000_000 diff --git a/providers/aimlapi/models/openai/gpt-4.1-nano.toml b/providers/aimlapi/models/openai/gpt-4.1-nano.toml new file mode 100644 index 00000000000..dbc624c33b7 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-4.1-nano.toml @@ -0,0 +1,9 @@ +base_model = "openai/gpt-4.1-nano" + +[cost] +input = 0.13 +output = 0.52 +cache_read = 0.0325 + +[limit] +context = 1_000_000 diff --git a/providers/aimlapi/models/openai/gpt-4.1.toml b/providers/aimlapi/models/openai/gpt-4.1.toml index 75228bcff3b..e522294beb7 100644 --- a/providers/aimlapi/models/openai/gpt-4.1.toml +++ b/providers/aimlapi/models/openai/gpt-4.1.toml @@ -1,24 +1,6 @@ -name = "GPT-4.1" -description = "GPT-4.1 is OpenAI's large-context language model with up to 1M token input and improved instruction following." -family = "gpt" -release_date = "2025-08-07" -last_updated = "2025-08-07" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "openai/gpt-4.1" [cost] input = 2.6 output = 10.4 cache_read = 0.65 - -[limit] -context = 1_047_576 -input = 1_047_576 -output = 32_768 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-4.toml b/providers/aimlapi/models/openai/gpt-4.toml new file mode 100644 index 00000000000..2977fbeb28b --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-4.toml @@ -0,0 +1,9 @@ +base_model = "openai/gpt-4" + +[cost] +input = 39 +output = 78 +cache_read = 39 + +[limit] +context = 8_000 diff --git a/providers/aimlapi/models/openai/gpt-4o-2024-05-13.toml b/providers/aimlapi/models/openai/gpt-4o-2024-05-13.toml index faa36078832..fbfb5174ab3 100644 --- a/providers/aimlapi/models/openai/gpt-4o-2024-05-13.toml +++ b/providers/aimlapi/models/openai/gpt-4o-2024-05-13.toml @@ -1,23 +1,8 @@ -name = "GPT-4o" -description = "Multimodal AI model by OpenAI enhancing human-computer interaction." -family = "gpt" -release_date = "2025-08-07" -last_updated = "2025-08-07" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "openai/gpt-4o-2024-05-13" [cost] input = 6.5 output = 19.5 [limit] -context = 128_000 -input = 128_000 output = 16_384 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-4o-2024-08-06.toml b/providers/aimlapi/models/openai/gpt-4o-2024-08-06.toml index 0e8522113e3..37cdb080980 100644 --- a/providers/aimlapi/models/openai/gpt-4o-2024-08-06.toml +++ b/providers/aimlapi/models/openai/gpt-4o-2024-08-06.toml @@ -1,24 +1,6 @@ -name = "GPT-4o" -description = "Multimodal AI model by OpenAI enhancing human-computer interaction." -family = "gpt" -release_date = "2025-08-07" -last_updated = "2025-08-07" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "openai/gpt-4o-2024-08-06" [cost] input = 3.25 output = 13 cache_read = 1.625 - -[limit] -context = 128_000 -input = 128_000 -output = 16_384 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-4o-2024-11-20.toml b/providers/aimlapi/models/openai/gpt-4o-2024-11-20.toml index 0e8522113e3..909b9284067 100644 --- a/providers/aimlapi/models/openai/gpt-4o-2024-11-20.toml +++ b/providers/aimlapi/models/openai/gpt-4o-2024-11-20.toml @@ -1,24 +1,6 @@ -name = "GPT-4o" -description = "Multimodal AI model by OpenAI enhancing human-computer interaction." -family = "gpt" -release_date = "2025-08-07" -last_updated = "2025-08-07" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "openai/gpt-4o-2024-11-20" [cost] input = 3.25 output = 13 cache_read = 1.625 - -[limit] -context = 128_000 -input = 128_000 -output = 16_384 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-4o-mini.toml b/providers/aimlapi/models/openai/gpt-4o-mini.toml new file mode 100644 index 00000000000..b987fdaeb95 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-4o-mini.toml @@ -0,0 +1,6 @@ +base_model = "openai/gpt-4o-mini" + +[cost] +input = 0.195 +output = 0.78 +cache_read = 0.0975 diff --git a/providers/aimlapi/models/openai/gpt-4o.toml b/providers/aimlapi/models/openai/gpt-4o.toml index 0e8522113e3..88760223978 100644 --- a/providers/aimlapi/models/openai/gpt-4o.toml +++ b/providers/aimlapi/models/openai/gpt-4o.toml @@ -1,24 +1,6 @@ -name = "GPT-4o" -description = "Multimodal AI model by OpenAI enhancing human-computer interaction." -family = "gpt" -release_date = "2025-08-07" -last_updated = "2025-08-07" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "openai/gpt-4o" [cost] input = 3.25 output = 13 cache_read = 1.625 - -[limit] -context = 128_000 -input = 128_000 -output = 16_384 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5-1.toml b/providers/aimlapi/models/openai/gpt-5-1.toml deleted file mode 100644 index 16a1f12c4b2..00000000000 --- a/providers/aimlapi/models/openai/gpt-5-1.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GPT-5.1" -description = "GPT-5.1 is an incremental update to GPT-5 with improved accuracy and instruction adherence." -family = "gpt" -release_date = "2025-11-14" -last_updated = "2025-11-14" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.625 -output = 13 -cache_read = 0.1625 - -[limit] -context = 400_000 -input = 400_000 -output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5-2-chat-latest.toml b/providers/aimlapi/models/openai/gpt-5-2-chat-latest.toml deleted file mode 100644 index cc5e3c7f319..00000000000 --- a/providers/aimlapi/models/openai/gpt-5-2-chat-latest.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GPT-5.2 Chat Latest" -description = "The latest conversational release in the GPT-5.2 series with improved dialogue quality." -family = "gpt" -release_date = "2025-12-11" -last_updated = "2025-12-11" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 2.275 -output = 18.2 -cache_read = 0.2275 - -[limit] -context = 128_000 -input = 128_000 -output = 16_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5-2.toml b/providers/aimlapi/models/openai/gpt-5-2.toml deleted file mode 100644 index 127a4d72777..00000000000 --- a/providers/aimlapi/models/openai/gpt-5-2.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GPT-5.2" -description = "GPT-5.2 is OpenAI's next-generation language model with extended context and enhanced performance." -family = "gpt" -release_date = "2025-12-11" -last_updated = "2025-12-11" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 2.275 -output = 18.2 -cache_read = 0.2275 - -[limit] -context = 400_000 -input = 400_000 -output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5-2025-08-07.toml b/providers/aimlapi/models/openai/gpt-5-2025-08-07.toml deleted file mode 100644 index 23a961267c8..00000000000 --- a/providers/aimlapi/models/openai/gpt-5-2025-08-07.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GPT-5" -description = "GPT-5 is OpenAI's most advanced language model with a 400K token context window and superior reasoning capabilities." -family = "gpt" -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.625 -output = 13 -cache_read = 0.1625 - -[limit] -context = 400_000 -input = 400_000 -output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5-mini-2025-08-07.toml b/providers/aimlapi/models/openai/gpt-5-mini-2025-08-07.toml deleted file mode 100644 index f2fe2f9dd7c..00000000000 --- a/providers/aimlapi/models/openai/gpt-5-mini-2025-08-07.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GPT-5 Mini" -description = "GPT-5 mini delivers GPT-5 capabilities in a smaller, faster package optimized for cost-sensitive applications." -family = "gpt" -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.325 -output = 2.6 -cache_read = 0.0325 - -[limit] -context = 400_000 -input = 400_000 -output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5-mini.toml b/providers/aimlapi/models/openai/gpt-5-mini.toml index 7ffe48a8ab9..69fae50bb44 100644 --- a/providers/aimlapi/models/openai/gpt-5-mini.toml +++ b/providers/aimlapi/models/openai/gpt-5-mini.toml @@ -1,24 +1,10 @@ -name = "GPT-5 Mini" -description = "GPT-5 mini delivers GPT-5 capabilities in a smaller, faster package optimized for cost-sensitive applications." -family = "gpt" -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "openai/gpt-5-mini" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.378235 output = 3.02588 cache_read = 0.34385 - -[limit] -context = 400_000 -input = 400_000 -output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5-nano-2025-08-07.toml b/providers/aimlapi/models/openai/gpt-5-nano-2025-08-07.toml deleted file mode 100644 index 82f8edb5710..00000000000 --- a/providers/aimlapi/models/openai/gpt-5-nano-2025-08-07.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GPT-5 Nano" -description = "GPT-5 nano is the most lightweight GPT-5 variant, built for ultra-low latency use cases." -family = "gpt" -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.065 -output = 0.52 -cache_read = 0.0065 - -[limit] -context = 400_000 -input = 400_000 -output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5-nano.toml b/providers/aimlapi/models/openai/gpt-5-nano.toml index 6680551d190..56bebab6927 100644 --- a/providers/aimlapi/models/openai/gpt-5-nano.toml +++ b/providers/aimlapi/models/openai/gpt-5-nano.toml @@ -1,24 +1,10 @@ -name = "GPT-5 Nano" -description = "GPT-5 nano is the most lightweight GPT-5 variant, built for ultra-low latency use cases." -family = "gpt" -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "openai/gpt-5-nano" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.075647 output = 0.605176 cache_read = 0.06877 - -[limit] -context = 400_000 -input = 400_000 -output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5-pro.toml b/providers/aimlapi/models/openai/gpt-5-pro.toml index d00e8536048..ee8cb7ee067 100644 --- a/providers/aimlapi/models/openai/gpt-5-pro.toml +++ b/providers/aimlapi/models/openai/gpt-5-pro.toml @@ -1,23 +1,9 @@ -name = "GPT-5 Pro" -description = "GPT-5 Pro is the most powerful GPT-5 variant, offering extended output and superior performance for complex tasks." -family = "gpt" -release_date = "2025-11-21" -last_updated = "2025-11-21" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "openai/gpt-5-pro" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 20.631 output = 165.048 - -[limit] -context = 400_000 -input = 400_000 -output = 272_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5.1-2025-11-13.toml b/providers/aimlapi/models/openai/gpt-5.1-2025-11-13.toml deleted file mode 100644 index 16a1f12c4b2..00000000000 --- a/providers/aimlapi/models/openai/gpt-5.1-2025-11-13.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GPT-5.1" -description = "GPT-5.1 is an incremental update to GPT-5 with improved accuracy and instruction adherence." -family = "gpt" -release_date = "2025-11-14" -last_updated = "2025-11-14" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.625 -output = 13 -cache_read = 0.1625 - -[limit] -context = 400_000 -input = 400_000 -output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5.1-codex-max.toml b/providers/aimlapi/models/openai/gpt-5.1-codex-max.toml new file mode 100644 index 00000000000..d6258f1209a --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.1-codex-max.toml @@ -0,0 +1,10 @@ +base_model = "openai/gpt-5.1-codex-max" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 3.4385 +output = 27.508 +cache_read = 0.34385 diff --git a/providers/aimlapi/models/openai/gpt-5.1-codex-mini.toml b/providers/aimlapi/models/openai/gpt-5.1-codex-mini.toml index 07247594f82..fa3cd2062d2 100644 --- a/providers/aimlapi/models/openai/gpt-5.1-codex-mini.toml +++ b/providers/aimlapi/models/openai/gpt-5.1-codex-mini.toml @@ -1,24 +1,10 @@ -name = "GPT-5.1 Codex Mini" -description = "GPT-5.1 Codex Mini is a lighter coding model optimized for fast code completions and suggestions." -family = "gpt" -release_date = "2025-11-14" -last_updated = "2025-11-14" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "openai/gpt-5.1-codex-mini" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.34385 output = 2.7508 cache_read = 0.041262 - -[limit] -context = 400_000 -input = 400_000 -output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5.1-codex.toml b/providers/aimlapi/models/openai/gpt-5.1-codex.toml index a4e5580499c..a46ce9b3fde 100644 --- a/providers/aimlapi/models/openai/gpt-5.1-codex.toml +++ b/providers/aimlapi/models/openai/gpt-5.1-codex.toml @@ -1,24 +1,10 @@ -name = "GPT-5.1 Codex" -description = "GPT-5.1 Codex is a coding-specialized model from the GPT-5.1 series for code generation and understanding." -family = "gpt" -release_date = "2025-11-14" -last_updated = "2025-11-14" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "openai/gpt-5.1-codex" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 3.4385 output = 27.508 cache_read = 0.34385 - -[limit] -context = 400_000 -input = 400_000 -output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5.2-2025-12-11.toml b/providers/aimlapi/models/openai/gpt-5.2-2025-12-11.toml deleted file mode 100644 index 127a4d72777..00000000000 --- a/providers/aimlapi/models/openai/gpt-5.2-2025-12-11.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GPT-5.2" -description = "GPT-5.2 is OpenAI's next-generation language model with extended context and enhanced performance." -family = "gpt" -release_date = "2025-12-11" -last_updated = "2025-12-11" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 2.275 -output = 18.2 -cache_read = 0.2275 - -[limit] -context = 400_000 -input = 400_000 -output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5.2-chat-latest.toml b/providers/aimlapi/models/openai/gpt-5.2-chat-latest.toml deleted file mode 100644 index cc5e3c7f319..00000000000 --- a/providers/aimlapi/models/openai/gpt-5.2-chat-latest.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GPT-5.2 Chat Latest" -description = "The latest conversational release in the GPT-5.2 series with improved dialogue quality." -family = "gpt" -release_date = "2025-12-11" -last_updated = "2025-12-11" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 2.275 -output = 18.2 -cache_read = 0.2275 - -[limit] -context = 128_000 -input = 128_000 -output = 16_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5.2-codex.toml b/providers/aimlapi/models/openai/gpt-5.2-codex.toml index 57ce8a12896..90bdfd92974 100644 --- a/providers/aimlapi/models/openai/gpt-5.2-codex.toml +++ b/providers/aimlapi/models/openai/gpt-5.2-codex.toml @@ -1,24 +1,10 @@ -name = "GPT-5.2 Codex" -description = "GPT-5.2 Codex is a coding-specialized model from the GPT-5.2 series for advanced code generation." -family = "gpt" -release_date = "2026-01-21" -last_updated = "2026-01-21" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "openai/gpt-5.2-codex" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 4.8139 output = 38.5112 cache_read = 0.48139 - -[limit] -context = 400_000 -input = 400_000 -output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5.2-pro.toml b/providers/aimlapi/models/openai/gpt-5.2-pro.toml index dcdfb431f96..bf69a62b5f1 100644 --- a/providers/aimlapi/models/openai/gpt-5.2-pro.toml +++ b/providers/aimlapi/models/openai/gpt-5.2-pro.toml @@ -1,23 +1,9 @@ -name = "GPT-5.2 Pro" -description = "GPT-5.2 Pro is the flagship GPT-5.2 variant offering maximum output size and reasoning depth." -family = "gpt" -release_date = "2025-12-11" -last_updated = "2025-12-11" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "openai/gpt-5.2-pro" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 28.8834 output = 231.0672 - -[limit] -context = 400_000 -input = 400_000 -output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5.3-codex.toml b/providers/aimlapi/models/openai/gpt-5.3-codex.toml index dfe1a91d921..f69e9a098bf 100644 --- a/providers/aimlapi/models/openai/gpt-5.3-codex.toml +++ b/providers/aimlapi/models/openai/gpt-5.3-codex.toml @@ -1,24 +1,10 @@ -name = "GPT-5.3 Codex" -description = "GPT-5.3 Codex is a coding-specialized model from the GPT-5.3 series for advanced code generation." -family = "gpt" -release_date = "2026-03-02" -last_updated = "2026-03-02" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "openai/gpt-5.3-codex" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 4.8139 output = 38.5112 cache_read = 0.48139 - -[limit] -context = 400_000 -input = 400_000 -output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-5.4-mini.toml b/providers/aimlapi/models/openai/gpt-5.4-mini.toml new file mode 100644 index 00000000000..4bc5a1e2e7f --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.4-mini.toml @@ -0,0 +1,10 @@ +base_model = "openai/gpt-5.4-mini" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 2.0631 +output = 12.3786 +cache_read = 0.20631 diff --git a/providers/aimlapi/models/openai/gpt-5.4-nano.toml b/providers/aimlapi/models/openai/gpt-5.4-nano.toml new file mode 100644 index 00000000000..f746e7bfc86 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.4-nano.toml @@ -0,0 +1,10 @@ +base_model = "openai/gpt-5.4-nano" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.27508 +output = 1.71925 +cache_read = 0.027508 diff --git a/providers/aimlapi/models/openai/gpt-5.4-pro.toml b/providers/aimlapi/models/openai/gpt-5.4-pro.toml new file mode 100644 index 00000000000..8995a0fc847 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.4-pro.toml @@ -0,0 +1,9 @@ +base_model = "openai/gpt-5.4-pro" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 41.262 +output = 247.572 diff --git a/providers/aimlapi/models/openai/gpt-5.5-pro.toml b/providers/aimlapi/models/openai/gpt-5.5-pro.toml new file mode 100644 index 00000000000..3e419c38335 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.5-pro.toml @@ -0,0 +1,9 @@ +base_model = "openai/gpt-5.5-pro" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 41.262 +output = 247.572 diff --git a/providers/aimlapi/models/openai/gpt-5.6-luna-pro.toml b/providers/aimlapi/models/openai/gpt-5.6-luna-pro.toml new file mode 100644 index 00000000000..3e68a281b4f --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.6-luna-pro.toml @@ -0,0 +1,10 @@ +base_model = "openai/gpt-5.6-luna" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.27508 +output = 1.65048 +cache_read = 0.027508 diff --git a/providers/aimlapi/models/openai/gpt-5.6-luna.toml b/providers/aimlapi/models/openai/gpt-5.6-luna.toml new file mode 100644 index 00000000000..8334dc563ca --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.6-luna.toml @@ -0,0 +1,10 @@ +base_model = "openai/gpt-5.6-luna" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.26 +output = 1.56 +cache_read = 0.026 diff --git a/providers/aimlapi/models/openai/gpt-5.6-sol-pro.toml b/providers/aimlapi/models/openai/gpt-5.6-sol-pro.toml new file mode 100644 index 00000000000..42778ea6f24 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.6-sol-pro.toml @@ -0,0 +1,10 @@ +base_model = "openai/gpt-5.6-sol" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 5.5016 +output = 27.508 +cache_read = 0.55016 diff --git a/providers/aimlapi/models/openai/gpt-5.6-sol.toml b/providers/aimlapi/models/openai/gpt-5.6-sol.toml new file mode 100644 index 00000000000..ebe354795f9 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.6-sol.toml @@ -0,0 +1,10 @@ +base_model = "openai/gpt-5.6-sol" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 5.2 +output = 26 +cache_read = 0.52 diff --git a/providers/aimlapi/models/openai/gpt-5.6-terra-pro.toml b/providers/aimlapi/models/openai/gpt-5.6-terra-pro.toml new file mode 100644 index 00000000000..e5c75dbe409 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.6-terra-pro.toml @@ -0,0 +1,10 @@ +base_model = "openai/gpt-5.6-terra" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 2.7508 +output = 16.5048 +cache_read = 0.27508 diff --git a/providers/aimlapi/models/openai/gpt-5.6-terra.toml b/providers/aimlapi/models/openai/gpt-5.6-terra.toml new file mode 100644 index 00000000000..f12c3f71693 --- /dev/null +++ b/providers/aimlapi/models/openai/gpt-5.6-terra.toml @@ -0,0 +1,10 @@ +base_model = "openai/gpt-5.6-terra" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 2.6 +output = 15.6 +cache_read = 0.26 diff --git a/providers/aimlapi/models/openai/gpt-5.toml b/providers/aimlapi/models/openai/gpt-5.toml index 4244e5c135e..71b1e3b3138 100644 --- a/providers/aimlapi/models/openai/gpt-5.toml +++ b/providers/aimlapi/models/openai/gpt-5.toml @@ -1,24 +1,10 @@ -name = "GPT-5" -description = "GPT-5 is OpenAI's most advanced language model with a 400K token context window and superior reasoning capabilities." -family = "gpt" -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "openai/gpt-5" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 1.891175 output = 15.1294 cache_read = 1.71925 - -[limit] -context = 400_000 -input = 400_000 -output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-oss-120b.toml b/providers/aimlapi/models/openai/gpt-oss-120b.toml index 3e92a8894ef..55a8f241e13 100644 --- a/providers/aimlapi/models/openai/gpt-oss-120b.toml +++ b/providers/aimlapi/models/openai/gpt-oss-120b.toml @@ -1,13 +1,8 @@ -name = "GPT OSS 120B" -description = "GPT OSS 120B is the most powerful Open AI open weight model." -family = "gpt-oss" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "openai/gpt-oss-120b" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.48139 @@ -16,9 +11,4 @@ cache_read = 0.48139 [limit] context = 131_000 -input = 131_000 output = 131_072 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/gpt-oss-20b.toml b/providers/aimlapi/models/openai/gpt-oss-20b.toml index 0fa6aded63b..68998c7a33e 100644 --- a/providers/aimlapi/models/openai/gpt-oss-20b.toml +++ b/providers/aimlapi/models/openai/gpt-oss-20b.toml @@ -1,13 +1,8 @@ -name = "GPT OSS 20B" -description = "GPT OSS 20B is a powerful Open AI open weight model." -family = "gpt-oss" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "openai/gpt-oss-20b" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.103155 @@ -16,9 +11,4 @@ cache_read = 0.0515775 [limit] context = 131_000 -input = 131_000 output = 131_072 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/o1-pro.toml b/providers/aimlapi/models/openai/o1-pro.toml new file mode 100644 index 00000000000..aee7f3e2ce5 --- /dev/null +++ b/providers/aimlapi/models/openai/o1-pro.toml @@ -0,0 +1,9 @@ +base_model = "openai/o1-pro" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 206.31 +output = 825.24 diff --git a/providers/aimlapi/models/openai/o1.toml b/providers/aimlapi/models/openai/o1.toml new file mode 100644 index 00000000000..7704addd34a --- /dev/null +++ b/providers/aimlapi/models/openai/o1.toml @@ -0,0 +1,13 @@ +base_model = "openai/o1" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 19.5 +output = 78 +cache_read = 9.75 + +[limit] +context = 128_000 diff --git a/providers/aimlapi/models/openai/o3-2025-04-16.toml b/providers/aimlapi/models/openai/o3-2025-04-16.toml deleted file mode 100644 index 366eb3bdea7..00000000000 --- a/providers/aimlapi/models/openai/o3-2025-04-16.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "o3" -description = "OpenAI o3 is a frontier reasoning model with exceptional performance on complex scientific, math, and coding tasks." -family = "o" -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 2.6 -output = 10.4 -cache_read = 0.65 - -[limit] -context = 200_000 -input = 200_000 -output = 100_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/openai/o3-mini.toml b/providers/aimlapi/models/openai/o3-mini.toml new file mode 100644 index 00000000000..2f7f1df6a8a --- /dev/null +++ b/providers/aimlapi/models/openai/o3-mini.toml @@ -0,0 +1,10 @@ +base_model = "openai/o3-mini" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 1.43 +output = 5.72 +cache_read = 0.715 diff --git a/providers/aimlapi/models/openai/o3-pro.toml b/providers/aimlapi/models/openai/o3-pro.toml new file mode 100644 index 00000000000..20e80f25699 --- /dev/null +++ b/providers/aimlapi/models/openai/o3-pro.toml @@ -0,0 +1,9 @@ +base_model = "openai/o3-pro" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 27.508 +output = 110.032 diff --git a/providers/aimlapi/models/openai/o4-mini-2025-04-16.toml b/providers/aimlapi/models/openai/o4-mini-2025-04-16.toml deleted file mode 100644 index 399abd13e1d..00000000000 --- a/providers/aimlapi/models/openai/o4-mini-2025-04-16.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "o4 Mini" -description = "OpenAI o4-mini is a compact reasoning model that delivers strong performance with lower latency and cost." -family = "o" -release_date = "2025-08-07" -last_updated = "2025-08-07" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.43 -output = 5.72 -cache_read = 0.3575 - -[limit] -context = 200_000 -input = 200_000 -output = 100_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/perplexity/sonar-pro.toml b/providers/aimlapi/models/perplexity/sonar-pro.toml index 5d88a2f9ea4..d353019ac3a 100644 --- a/providers/aimlapi/models/perplexity/sonar-pro.toml +++ b/providers/aimlapi/models/perplexity/sonar-pro.toml @@ -1,23 +1,8 @@ -name = "Sonar Pro" -description = "Perplexity Sonar Pro is an advanced search-augmented language model with deeper reasoning and broader web coverage." -family = "sonar-pro" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "perplexity/sonar-pro" [cost] input = 3.9 output = 19.5 [limit] -context = 200_000 -input = 200_000 output = 100_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/perplexity/sonar.toml b/providers/aimlapi/models/perplexity/sonar.toml index 036db6a1d8d..8b062038318 100644 --- a/providers/aimlapi/models/perplexity/sonar.toml +++ b/providers/aimlapi/models/perplexity/sonar.toml @@ -1,23 +1,8 @@ -name = "Sonar" -description = "Perplexity Sonar is a fast search-augmented language model grounded in real-time web data for up-to-date answers." -family = "sonar" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "perplexity/sonar" [cost] input = 1.3 output = 1.3 [limit] -context = 128_000 -input = 128_000 output = 100_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/poolside/laguna-s-2.1.toml b/providers/aimlapi/models/poolside/laguna-s-2.1.toml new file mode 100644 index 00000000000..7bbd01df825 --- /dev/null +++ b/providers/aimlapi/models/poolside/laguna-s-2.1.toml @@ -0,0 +1,13 @@ +base_model = "poolside/laguna-s-2.1" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.123786 +output = 0.247572 +cache_read = 0.0082524 + +[limit] +output = 131_072 diff --git a/providers/aimlapi/models/poolside/laguna-xs-2.1.toml b/providers/aimlapi/models/poolside/laguna-xs-2.1.toml new file mode 100644 index 00000000000..011a014a3c3 --- /dev/null +++ b/providers/aimlapi/models/poolside/laguna-xs-2.1.toml @@ -0,0 +1,10 @@ +base_model = "poolside/laguna-xs-2.1" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.082524 +output = 0.165048 +cache_read = 0.041262 diff --git a/providers/aimlapi/models/sakana/fugu-ultra.toml b/providers/aimlapi/models/sakana/fugu-ultra.toml index 37ecd1d9ae2..f14cb9af343 100644 --- a/providers/aimlapi/models/sakana/fugu-ultra.toml +++ b/providers/aimlapi/models/sakana/fugu-ultra.toml @@ -1,13 +1,8 @@ -name = "Fugu Ultra" -description = "Fugu Ultra is the higher-performance model in Sakana AI's Fugu family. Rather than a standard monolithic language model, it is a learned multi-agent orchestration system: a model trained to route tasks across a swappable pool of underlying models and to recursively call instances of itself. It is optimized for complex, multi-step reasoning, coding, and agentic workflows, with orchestration tokens billed as standard input/output tokens." -family = "fugu" -release_date = "2026-06-24" -last_updated = "2026-06-24" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "sakana/fugu-ultra" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 6.877 @@ -15,10 +10,4 @@ output = 41.262 cache_read = 0.6877 [limit] -context = 1_000_000 -input = 1_000_000 output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/sakana/sakana-namazu.toml b/providers/aimlapi/models/sakana/sakana-namazu.toml index 1aae96e483d..0910a71e8c8 100644 --- a/providers/aimlapi/models/sakana/sakana-namazu.toml +++ b/providers/aimlapi/models/sakana/sakana-namazu.toml @@ -1,24 +1,10 @@ -name = "Sakana Namazu" -description = "Sakana Namazu is a Japanese-specialized reasoning model from Sakana AI, based on Kimi K2.6 with additional training for Japanese language and business contexts. It is suited for Japanese instruction following and business workflows, supports a 262K context window, and accepts text, image, and file inputs." -family = "sakana-namazu" -release_date = "2026-08-11" -last_updated = "2026-08-11" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "sakana/sakana-namazu" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 1.30663 output = 5.5016 cache_read = 0.20631 - -[limit] -context = 262_144 -input = 262_144 -output = 65_536 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/stepfun/step-3.7-flash.toml b/providers/aimlapi/models/stepfun/step-3.7-flash.toml deleted file mode 100644 index 06c8efc9b26..00000000000 --- a/providers/aimlapi/models/stepfun/step-3.7-flash.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Step 3.7 Flash" -description = "Step 3.7 Flash is a multimodal Mixture-of-Experts model from StepFun with text, image, and video understanding. It supports a 256K context window and is optimized for fast inference, reasoning, agent workflows, and long-context productivity tasks." -family = "step" -release_date = "2026-06-01" -last_updated = "2026-06-01" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.27508 -output = 1.58171 -cache_read = 0.055016 - -[limit] -context = 256_000 -input = 256_000 -output = 256_000 - -[modalities] -input = ["image", "text", "video"] -output = ["text"] diff --git a/providers/aimlapi/models/tencent/hy-mt2-1.8b.toml b/providers/aimlapi/models/tencent/hy-mt2-1.8b.toml deleted file mode 100644 index 63fb3e2d26f..00000000000 --- a/providers/aimlapi/models/tencent/hy-mt2-1.8b.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Hy-MT2-1.8B" -description = "Hy-MT2-1.8B is a compact 1.8B-parameter translation model from Tencent. It supports 33 language pairs and five Chinese dialect and minority-language pairs, with workflows for structured, delimiter-based, contextual, glossary-based, and style-guided translation." -family = "Hy" -release_date = "2026-08-20" -last_updated = "2026-08-20" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.06051759999999999 -output = 0.2434458 - -[limit] -context = 8_192 -input = 8_192 -output = 4_096 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/tencent/hy-mt2-30b-a3b.toml b/providers/aimlapi/models/tencent/hy-mt2-30b-a3b.toml deleted file mode 100644 index c12eadca9b3..00000000000 --- a/providers/aimlapi/models/tencent/hy-mt2-30b-a3b.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Hy-MT2-30B-A3B" -description = "Hy-MT2-30B-A3B is Tencent's flagship translation model in the Hy-MT2 family. It supports 33 language pairs plus five Chinese dialect and minority-language pairs, with workflows for structured, delimiter-based, contextual and glossary-based translation." -family = "Hy" -release_date = "2026-08-20" -last_updated = "2026-08-20" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.1017796 -output = 0.405743 - -[limit] -context = 8_192 -input = 8_192 -output = 4_096 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/tencent/hy-mt2-7b.toml b/providers/aimlapi/models/tencent/hy-mt2-7b.toml deleted file mode 100644 index 8b2ab3a201f..00000000000 --- a/providers/aimlapi/models/tencent/hy-mt2-7b.toml +++ /dev/null @@ -1,23 +0,0 @@ -name = "Hy-MT2-7B" -description = "Hy-MT2-7B is a 7B-parameter translation model from Tencent. It supports 33 language pairs and five Chinese dialect and minority-language pairs, with workflows for structured, delimiter-based, contextual, glossary-based, and style-guided translation." -family = "Hy" -release_date = "2026-08-19" -last_updated = "2026-08-19" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.1017796 -output = 0.405743 - -[limit] -context = 8_192 -input = 8_192 -output = 4_096 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/tencent/hy3.toml b/providers/aimlapi/models/tencent/hy3.toml new file mode 100644 index 00000000000..a8d79e8ba93 --- /dev/null +++ b/providers/aimlapi/models/tencent/hy3.toml @@ -0,0 +1,14 @@ +base_model = "tencent/hy3" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.27508 +output = 1.10032 +cache_read = 0.06877 + +[limit] +context = 262_144 +output = 131_072 diff --git a/providers/aimlapi/models/tencent/hy4-preview.toml b/providers/aimlapi/models/tencent/hy4-preview.toml index e9ca1a5cb9d..cc65e8c53b8 100644 --- a/providers/aimlapi/models/tencent/hy4-preview.toml +++ b/providers/aimlapi/models/tencent/hy4-preview.toml @@ -1,12 +1,8 @@ -name = "Hy4 preview" -description = "Hy4 preview is a mixture-of-experts model from Tencent with 49B active parameters out of 770B total, built for coding agents, complex tool-use workflows and productivity tasks. It has a 1M-token context window and configurable reasoning effort." -release_date = "2026-08-28" -last_updated = "2026-08-28" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "tencent/hy4-preview" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 1.1470836 @@ -15,9 +11,3 @@ cache_read = 0.0577668 [limit] context = 1_048_576 -input = 1_048_576 -output = 64_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/thinkingmachines/inkling-small.toml b/providers/aimlapi/models/thinkingmachines/inkling-small.toml index 650b4a42afb..8ee54fbd811 100644 --- a/providers/aimlapi/models/thinkingmachines/inkling-small.toml +++ b/providers/aimlapi/models/thinkingmachines/inkling-small.toml @@ -1,12 +1,8 @@ -name = "Inkling Small" -description = "Inkling Small is an open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 12B active parameters out of 276B total. It is positioned as the smaller, more efficient member of the Inkling family." -release_date = "2026-07-30" -last_updated = "2026-07-30" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "thinkingmachines/inkling-small" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.797732 @@ -15,9 +11,4 @@ cache_read = 0.1595464 [limit] context = 524_288 -input = 524_288 output = 262_144 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/thinkingmachines/inkling.toml b/providers/aimlapi/models/thinkingmachines/inkling.toml new file mode 100644 index 00000000000..b7eb76e80e5 --- /dev/null +++ b/providers/aimlapi/models/thinkingmachines/inkling.toml @@ -0,0 +1,10 @@ +base_model = "thinkingmachines/inkling" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 1.3754 +output = 5.57037 +cache_read = 0.233818 diff --git a/providers/aimlapi/models/upstage/solar-pro4.toml b/providers/aimlapi/models/upstage/solar-pro4.toml index 64980444a79..ac67690de8d 100644 --- a/providers/aimlapi/models/upstage/solar-pro4.toml +++ b/providers/aimlapi/models/upstage/solar-pro4.toml @@ -1,24 +1,10 @@ -name = "Solar Pro 4" -description = "Solar Pro 4 is a large language model from Upstage. It is suited for agentic workflows, office productivity, document-intensive work, and coding." -family = "solar" -release_date = "2026-08-10" -last_updated = "2026-08-10" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "upstage/solar-pro4" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.041262 output = 0.165048 cache_read = 0.0082524 - -[limit] -context = 524_288 -input = 524_288 -output = 131_072 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-3-beta.toml b/providers/aimlapi/models/x-ai/grok-3-beta.toml deleted file mode 100644 index 9a3320db9b8..00000000000 --- a/providers/aimlapi/models/x-ai/grok-3-beta.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Grok 3 Beta" -description = "Grok 3 Beta is xAI's advanced language model with strong reasoning and broad knowledge capabilities." -family = "grok" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.625 -output = 3.25 -cache_read = 0.26 - -[limit] -context = 131_000 -input = 131_000 -output = 130_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-3-mini-beta.toml b/providers/aimlapi/models/x-ai/grok-3-mini-beta.toml deleted file mode 100644 index 0d390c78b7f..00000000000 --- a/providers/aimlapi/models/x-ai/grok-3-mini-beta.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Grok 3 Beta Mini" -description = "Grok 3 Beta Mini is a compact, efficient variant of Grok 3 optimized for speed and cost-effective inference." -family = "grok" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.625 -output = 3.25 -cache_read = 0.26 - -[limit] -context = 131_000 -input = 131_000 -output = 130_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-4-1-fast-non-reasoning.toml b/providers/aimlapi/models/x-ai/grok-4-1-fast-non-reasoning.toml deleted file mode 100644 index 8c6dd4465e2..00000000000 --- a/providers/aimlapi/models/x-ai/grok-4-1-fast-non-reasoning.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Grok 4.1 Fast Non-Reasoning" -description = "Grok 4.1 Fast without reasoning for faster responses and lower latency in high-performance agentic applications." -family = "grok" -release_date = "2025-11-20" -last_updated = "2025-11-20" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.625 -output = 3.25 -cache_read = 0.26 - -[limit] -context = 2_000_000 -input = 2_000_000 -output = 1_999_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-4-1-fast-reasoning.toml b/providers/aimlapi/models/x-ai/grok-4-1-fast-reasoning.toml deleted file mode 100644 index c3a03b76a6a..00000000000 --- a/providers/aimlapi/models/x-ai/grok-4-1-fast-reasoning.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Grok 4.1 Fast Reasoning" -description = "Grok 4.1 Fast is a frontier multimodal model optimized for high-performance agentic tool calling with advanced reasoning capabilities." -family = "grok" -release_date = "2025-11-20" -last_updated = "2025-11-20" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.625 -output = 3.25 -cache_read = 0.26 - -[limit] -context = 2_000_000 -input = 2_000_000 -output = 1_999_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-4-20-0309-non-reasoning.toml b/providers/aimlapi/models/x-ai/grok-4-20-0309-non-reasoning.toml deleted file mode 100644 index 79ee2f9180c..00000000000 --- a/providers/aimlapi/models/x-ai/grok-4-20-0309-non-reasoning.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Grok 4.20 Non-Reasoning" -description = "Grok 4.20 Non-Reasoning — a general-purpose language model optimized for standard text generation and conversational tasks, without extended reasoning capabilities." -family = "grok" -release_date = "2026-04-09" -last_updated = "2026-04-09" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.625 -output = 3.25 -cache_read = 0.26 - -[limit] -context = 2_000_000 -input = 2_000_000 -output = 1_999_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-4-20-0309-reasoning.toml b/providers/aimlapi/models/x-ai/grok-4-20-0309-reasoning.toml deleted file mode 100644 index bb2642d914b..00000000000 --- a/providers/aimlapi/models/x-ai/grok-4-20-0309-reasoning.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Grok 4.20 Reasoning" -description = "Grok 4.20 Reasoning — a large language model designed for advanced reasoning tasks. It excels at multi-step problem solving, logical analysis, and complex instruction following." -family = "grok" -release_date = "2026-04-09" -last_updated = "2026-04-09" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.625 -output = 3.25 -cache_read = 0.26 - -[limit] -context = 2_000_000 -input = 2_000_000 -output = 1_999_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-4-3.toml b/providers/aimlapi/models/x-ai/grok-4-3.toml deleted file mode 100644 index 115e62d2432..00000000000 --- a/providers/aimlapi/models/x-ai/grok-4-3.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Grok 4.3" -description = "Grok 4.3 is a large language model by xAI for text generation and reasoning tasks. Supports up to 1M context tokens and is optimized for conversational AI and analytical workflows." -family = "grok" -release_date = "2026-05-05" -last_updated = "2026-05-05" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.625 -output = 3.25 -cache_read = 0.26 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 1_000_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-4-fast-non-reasoning.toml b/providers/aimlapi/models/x-ai/grok-4-fast-non-reasoning.toml deleted file mode 100644 index 7364af4b893..00000000000 --- a/providers/aimlapi/models/x-ai/grok-4-fast-non-reasoning.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Grok 4 Fast Non-Reasoning" -description = "Grok 4 Fast without reasoning for faster responses and lower latency." -family = "grok" -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.625 -output = 3.25 -cache_read = 0.26 - -[limit] -context = 2_000_000 -input = 2_000_000 -output = 1_999_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-4-fast-reasoning.toml b/providers/aimlapi/models/x-ai/grok-4-fast-reasoning.toml deleted file mode 100644 index 016fec90f40..00000000000 --- a/providers/aimlapi/models/x-ai/grok-4-fast-reasoning.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Grok 4 Fast Reasoning" -description = "Grok 4 Fast with always-on reasoning for complex problem-solving tasks." -family = "grok" -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.625 -output = 3.25 -cache_read = 0.26 - -[limit] -context = 2_000_000 -input = 2_000_000 -output = 1_999_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/x-ai/grok-code-fast-1.toml b/providers/aimlapi/models/x-ai/grok-code-fast-1.toml deleted file mode 100644 index 6ad387bf9bd..00000000000 --- a/providers/aimlapi/models/x-ai/grok-code-fast-1.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "Grok Code Fast 1" -description = "A speedy and economical reasoning model that excels at agentic coding with function calling and structured outputs." -family = "grok" -release_date = "2025-09-30" -last_updated = "2025-09-30" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.3 -output = 2.6 -cache_read = 0.26 - -[limit] -context = 256_000 -input = 256_000 -output = 255_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/xiaomi/mimo-v2.5-pro.toml b/providers/aimlapi/models/xiaomi/mimo-v2.5-pro.toml index 9586b604742..649a50bdb63 100644 --- a/providers/aimlapi/models/xiaomi/mimo-v2.5-pro.toml +++ b/providers/aimlapi/models/xiaomi/mimo-v2.5-pro.toml @@ -1,13 +1,8 @@ -name = "MiMo V2.5 Pro" -description = "MiMo-V2.5-Pro — a flagship reasoning model by Xiaomi designed for complex agentic workflows, long-horizon tasks, and advanced software engineering scenarios with a 1M token context window." -family = "mimo-v2.5-pro" -release_date = "2026-04-23" -last_updated = "2026-04-23" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "xiaomi/mimo-v2.5-pro" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.598299 @@ -16,9 +11,4 @@ cache_read = 0.00495144 [limit] context = 1_000_000 -input = 1_000_000 output = 128_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/xiaomi/mimo-v2.5.toml b/providers/aimlapi/models/xiaomi/mimo-v2.5.toml index cc34d0620fa..8ea2de6fccc 100644 --- a/providers/aimlapi/models/xiaomi/mimo-v2.5.toml +++ b/providers/aimlapi/models/xiaomi/mimo-v2.5.toml @@ -1,13 +1,8 @@ -name = "MiMo V2.5" -description = "MiMo-V2.5 — a native omnimodal model by Xiaomi for text, image, audio, and video understanding. It delivers strong agentic performance at lower cost, making it suitable for general-purpose AI applications." -family = "mimo-v2.5" -release_date = "2026-04-23" -last_updated = "2026-04-23" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false +base_model = "xiaomi/mimo-v2.5" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] [cost] input = 0.192556 @@ -16,9 +11,4 @@ cache_read = 0.0038511200000000004 [limit] context = 1_000_000 -input = 1_000_000 output = 128_000 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/z-ai/glm-4.5v.toml b/providers/aimlapi/models/z-ai/glm-4.5v.toml new file mode 100644 index 00000000000..c745623a2ad --- /dev/null +++ b/providers/aimlapi/models/z-ai/glm-4.5v.toml @@ -0,0 +1,13 @@ +base_model = "zhipuai/glm-4.5v" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.82524 +output = 2.47572 +cache_read = 0.151294 + +[limit] +context = 65_536 diff --git a/providers/aimlapi/models/z-ai/glm-4.6v.toml b/providers/aimlapi/models/z-ai/glm-4.6v.toml new file mode 100644 index 00000000000..ba8f70f7db0 --- /dev/null +++ b/providers/aimlapi/models/z-ai/glm-4.6v.toml @@ -0,0 +1,13 @@ +base_model = "zhipuai/glm-4.6v" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.41262 +output = 1.23786 +cache_read = 0.075647 + +[limit] +context = 131_072 diff --git a/providers/aimlapi/models/z-ai/glm-4.7-flash.toml b/providers/aimlapi/models/z-ai/glm-4.7-flash.toml new file mode 100644 index 00000000000..d828cf1badf --- /dev/null +++ b/providers/aimlapi/models/z-ai/glm-4.7-flash.toml @@ -0,0 +1,14 @@ +base_model = "zhipuai/glm-4.7-flash" + +[[reasoning_options]] +type = "effort" +values = ["none", "low", "medium", "high"] + +[cost] +input = 0.171925 +output = 0.6877 +cache_read = 0.013754 + +[limit] +context = 202_752 +output = 16_384 diff --git a/providers/aimlapi/models/z-ai/glm-5.3-flash.toml b/providers/aimlapi/models/z-ai/glm-5.3-flash.toml deleted file mode 100644 index 55777da55b8..00000000000 --- a/providers/aimlapi/models/z-ai/glm-5.3-flash.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GLM 5.3 Flash" -description = "GLM-5.3 Flash is the low-cost, high-throughput variant of Zhipu AI GLM-5.3, with a 1M-token context window and text, image and video understanding." -family = "glm" -release_date = "2026-08-26" -last_updated = "2026-08-26" -attachment = true -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.0975 -output = 0.325 -cache_read = 0.0195 - -[limit] -context = 1_048_576 -input = 1_048_576 -output = 131_072 - -[modalities] -input = ["image", "text"] -output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-4.5-air.toml b/providers/aimlapi/models/zhipu/glm-4.5-air.toml deleted file mode 100644 index 70f286dae39..00000000000 --- a/providers/aimlapi/models/zhipu/glm-4.5-air.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GLM 4.5 Air" -description = "Lightweight version of GLM-4.5 for cost-effective applications" -family = "glm" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.26 -output = 1.43 -cache_read = 0.039 - -[limit] -context = 128_000 -input = 128_000 -output = 98_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-4.5.toml b/providers/aimlapi/models/zhipu/glm-4.5.toml deleted file mode 100644 index a0e264a7738..00000000000 --- a/providers/aimlapi/models/zhipu/glm-4.5.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GLM 4.5" -description = "Advanced version of GLM-4.5 with enhanced capabilities and web search integration" -family = "glm" -release_date = "2025-09-09" -last_updated = "2025-09-09" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.78 -output = 2.86 -cache_read = 0.143 - -[limit] -context = 128_000 -input = 128_000 -output = 98_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-4.6.toml b/providers/aimlapi/models/zhipu/glm-4.6.toml deleted file mode 100644 index f2dd20741b1..00000000000 --- a/providers/aimlapi/models/zhipu/glm-4.6.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GLM 4.6" -description = "Advanced version of GLM-4.6 with web search integration and enhanced capabilities" -family = "glm" -release_date = "2025-10-02" -last_updated = "2025-10-02" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.78 -output = 2.86 -cache_read = 0.143 - -[limit] -context = 200_000 -input = 200_000 -output = 128_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-4.7.toml b/providers/aimlapi/models/zhipu/glm-4.7.toml deleted file mode 100644 index 70394acd2a1..00000000000 --- a/providers/aimlapi/models/zhipu/glm-4.7.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GLM 4.7" -description = "GLM-4.7 is Zhipu AI flagship text LLM optimized for agentic coding and stable multi-step reasoning, supporting long-context workflows." -family = "glm" -release_date = "2025-12-23" -last_updated = "2025-12-23" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 0.78 -output = 2.86 -cache_read = 0.143 - -[limit] -context = 200_000 -input = 200_000 -output = 128_000 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-5-2-fast-preview.toml b/providers/aimlapi/models/zhipu/glm-5-2-fast-preview.toml deleted file mode 100644 index e6cf4b4c3bd..00000000000 --- a/providers/aimlapi/models/zhipu/glm-5-2-fast-preview.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GLM 5.2 Fast Preview" -description = "A latency-optimized GLM-5.2 preview with the same 1M-token context window and 1.5–2× higher output throughput for real-time chat, agents, and coding." -family = "glm" -release_date = "2026-07-10" -last_updated = "2026-07-10" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 4.55 -output = 14.3 -cache_read = 0.91 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 131_072 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-5-2.toml b/providers/aimlapi/models/zhipu/glm-5-2.toml deleted file mode 100644 index 25dfee5cbaf..00000000000 --- a/providers/aimlapi/models/zhipu/glm-5-2.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GLM 5.2" -description = "GLM-5.2 is Zhipu AI flagship MoE LLM with a 1M-token context window, optimized for agentic coding, tool use, reasoning, and long-context tasks." -family = "glm" -release_date = "2026-06-16" -last_updated = "2026-06-16" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.82 -output = 5.72 -cache_read = 0.338 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 131_072 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-5-3.toml b/providers/aimlapi/models/zhipu/glm-5-3.toml deleted file mode 100644 index 39ef4b76da2..00000000000 --- a/providers/aimlapi/models/zhipu/glm-5-3.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GLM 5.3" -description = "GLM-5.3 is Zhipu AI flagship MoE LLM with a 1M-token context window, built for advanced text generation and deep reasoning." -family = "glm" -release_date = "2026-08-17" -last_updated = "2026-08-17" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.82 -output = 5.72 -cache_read = 0.338 - -[limit] -context = 1_024_000 -input = 1_024_000 -output = 131_072 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-5.2.toml b/providers/aimlapi/models/zhipu/glm-5.2.toml deleted file mode 100644 index 25dfee5cbaf..00000000000 --- a/providers/aimlapi/models/zhipu/glm-5.2.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GLM 5.2" -description = "GLM-5.2 is Zhipu AI flagship MoE LLM with a 1M-token context window, optimized for agentic coding, tool use, reasoning, and long-context tasks." -family = "glm" -release_date = "2026-06-16" -last_updated = "2026-06-16" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.82 -output = 5.72 -cache_read = 0.338 - -[limit] -context = 1_000_000 -input = 1_000_000 -output = 131_072 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-5.3.toml b/providers/aimlapi/models/zhipu/glm-5.3.toml deleted file mode 100644 index 39ef4b76da2..00000000000 --- a/providers/aimlapi/models/zhipu/glm-5.3.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GLM 5.3" -description = "GLM-5.3 is Zhipu AI flagship MoE LLM with a 1M-token context window, built for advanced text generation and deep reasoning." -family = "glm" -release_date = "2026-08-17" -last_updated = "2026-08-17" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.82 -output = 5.72 -cache_read = 0.338 - -[limit] -context = 1_024_000 -input = 1_024_000 -output = 131_072 - -[modalities] -input = ["text"] -output = ["text"] diff --git a/providers/aimlapi/models/zhipu/glm-5.toml b/providers/aimlapi/models/zhipu/glm-5.toml deleted file mode 100644 index 2e6c89f2318..00000000000 --- a/providers/aimlapi/models/zhipu/glm-5.toml +++ /dev/null @@ -1,24 +0,0 @@ -name = "GLM 5" -description = "GLM-5 is Zhipu AI next-generation 745B MoE LLM for high-quality text generation and reasoning, optimized for chat, long-form content, and instruction-following tasks." -family = "glm" -release_date = "2026-02-12" -last_updated = "2026-02-12" -attachment = false -reasoning = false -tool_call = false -structured_output = false -open_weights = false - -[cost] -input = 1.3 -output = 4.16 -cache_read = 0.26 - -[limit] -context = 200_000 -input = 200_000 -output = 128_000 - -[modalities] -input = ["text"] -output = ["text"] From 35a3b94587a927073fe4ad7bdf3f5e92ddcb4033 Mon Sep 17 00:00:00 2001 From: Stan Date: Fri, 4 Sep 2026 18:56:24 +0500 Subject: [PATCH 3/3] fix(aimlapi): exclude image surfaces and regenerate reasoning options from per-model schemas MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Addresses the review on this PR. Two separate causes, one in this sync and one upstream in the host's API. Image surfaces. `isChatTextModel` judged each record on its own modalities, but the catalog lists an id once per endpoint type and the CHAT record of an image model claims text output: `google/gemini-2.5-flash-image` appears both as `openai/image-generations` with `output = ["image"]` and as `openai/chat-completions` with `output = ["text"]`, and the same holds for the `gemini-3-pro-image` and `gemini-3.1-flash-image` families. The answer is not in the record, so the whole response is indexed first and an id this host serves on any media surface is refused regardless of what its chat record claims. The six image entries this admitted are deleted; `deleteMissing: false` retains files that merely fall out of the source, which is why removing them takes an explicit delete. Reasoning options. The uniform `["none", "low", "medium", "high"]` was not invented here — it was read faithfully from a host schema that published one boilerplate enum for every model. The host has since corrected that, and its published schemas now carry seven distinct ladders. Re-running the sync therefore produces per-model values: `none` disappears from GPT-5, Qwen3.8 2.4T A95B and Kimi K2 Thinking, and two models that previously documented no control at all (`alibaba/qwen3.6-max-preview`, `z-ai/glm-5.3-flash`) now do. Known remaining gap, stated rather than papered over: the host's per-model schema endpoint still understates GPT-5, publishing `low/medium/high` where the model demonstrably accepts `minimal` as well. That is a fix on the host side, not here; until it lands, this sync can only publish what the per-model schema states. --- packages/core/src/sync/providers/aimlapi.ts | 39 ++++++++++++++++++- .../qwen3.6-max-preview.toml} | 6 +-- .../models/alibaba/qwen3.8-2.4t-a95b.toml | 2 +- .../arcee-ai/trinity-large-thinking.toml | 2 +- .../models/google/gemini-2.5-flash-image.toml | 10 ----- .../models/google/gemini-2.5-flash-lite.toml | 3 +- .../google/gemini-3-pro-image-preview.toml | 10 ----- .../models/google/gemini-3-pro-image.toml | 10 ----- .../gemini-3.1-flash-image-preview.toml | 13 ------- .../google/gemini-3.1-flash-lite-image.toml | 12 ------ .../models/google/gemini-3.6-flash.toml | 2 +- .../models/google/gemini-3.7-flash.toml | 2 +- .../aimlapi/models/google/gemma-4-31b-it.toml | 2 +- .../models/mistralai/mistral-nemo.toml | 2 +- .../models/moonshotai/kimi-k2-thinking.toml | 4 +- .../aimlapi/models/openai/gpt-5-mini.toml | 2 +- .../aimlapi/models/openai/gpt-5-nano.toml | 2 +- .../aimlapi/models/openai/gpt-5-pro.toml | 2 +- .../aimlapi/models/openai/gpt-5.1-codex.toml | 2 +- .../aimlapi/models/openai/gpt-5.2-codex.toml | 2 +- .../aimlapi/models/openai/gpt-5.2-pro.toml | 2 +- .../aimlapi/models/openai/gpt-5.4-pro.toml | 2 +- .../aimlapi/models/openai/gpt-5.5-pro.toml | 2 +- providers/aimlapi/models/openai/gpt-5.toml | 2 +- .../aimlapi/models/openai/gpt-oss-120b.toml | 4 +- .../aimlapi/models/openai/gpt-oss-20b.toml | 4 +- providers/aimlapi/models/openai/o1-pro.toml | 2 +- providers/aimlapi/models/openai/o1.toml | 2 +- providers/aimlapi/models/openai/o3-mini.toml | 2 +- providers/aimlapi/models/openai/o3-pro.toml | 2 +- providers/aimlapi/models/z-ai/glm-4.5v.toml | 2 +- providers/aimlapi/models/z-ai/glm-4.6v.toml | 2 +- .../aimlapi/models/z-ai/glm-4.7-flash.toml | 2 +- .../aimlapi/models/z-ai/glm-5.3-flash.toml | 13 +++++++ 34 files changed, 83 insertions(+), 89 deletions(-) rename providers/aimlapi/models/{google/gemini-3.1-flash-image.toml => alibaba/qwen3.6-max-preview.toml} (54%) delete mode 100644 providers/aimlapi/models/google/gemini-2.5-flash-image.toml delete mode 100644 providers/aimlapi/models/google/gemini-3-pro-image-preview.toml delete mode 100644 providers/aimlapi/models/google/gemini-3-pro-image.toml delete mode 100644 providers/aimlapi/models/google/gemini-3.1-flash-image-preview.toml delete mode 100644 providers/aimlapi/models/google/gemini-3.1-flash-lite-image.toml create mode 100644 providers/aimlapi/models/z-ai/glm-5.3-flash.toml diff --git a/packages/core/src/sync/providers/aimlapi.ts b/packages/core/src/sync/providers/aimlapi.ts index c575a607f18..d8c7b25c185 100755 --- a/packages/core/src/sync/providers/aimlapi.ts +++ b/packages/core/src/sync/providers/aimlapi.ts @@ -84,8 +84,41 @@ function normalizeModalities(values: readonly string[] | null | undefined): Moda return [...seen]; } +/** + * Ids this host also serves on a non-text surface. + * + * The catalog lists an id once per endpoint type, and the chat-surface record of + * an image model claims text output. Measured 2026-09-04: + * `google/gemini-2.5-flash-image` appears both as `openai/image-generations` + * with `output: ["image"]` and as `openai/chat-completions` with + * `output: ["text"]`; the same holds for the `gemini-3-pro-image` and + * `gemini-3.1-flash-image` families. Judging a record only by its own modalities + * therefore admits image generators into a chat catalog. + * + * An id this host serves as a media model is not a text-only chat model, whatever + * its chat record claims. Populated from the whole response before any record is + * judged, because the answer is not in the record itself. + */ +const mediaOutputIDs = new Set(); + +function indexMediaOutputs(models: readonly AimlapiModel[]): void { + mediaOutputIDs.clear(); + for (const model of models) { + const declared = model.modalities?.output ?? []; + // `normalizeModalities` treats an empty list as text, so an undeclared + // record must not be read as evidence of anything. + if (declared.length === 0) continue; + if (normalizeModalities(declared).some((modality) => modality !== "text")) { + mediaOutputIDs.add(model.id); + } + } +} + function isChatTextModel(model: AimlapiModel): boolean { if (model.type !== CHAT_COMPLETIONS_TYPE) return false; + // Cross-surface check first: the chat record of a media model does not admit + // to being one. + if (mediaOutputIDs.has(model.id)) return false; const output = normalizeModalities(model.modalities?.output); // A chat model whose output is not purely text is a media model riding the // chat protocol, and does not belong in a chat catalog. @@ -227,11 +260,15 @@ export const aimlapi = { } const raw = await response.json(); const parsed = AimlapiResponse.parse(raw); + indexMediaOutputs(parsed.data); await attachReasoningEffort(parsed.data); return parsed; }, parseModels(raw) { - return AimlapiResponse.parse(raw).data; + const models = AimlapiResponse.parse(raw).data; + // Replays parse a cached payload without going through fetchModels. + indexMediaOutputs(models); + return models; }, translateModel(model, context) { if (!isChatTextModel(model)) return undefined; diff --git a/providers/aimlapi/models/google/gemini-3.1-flash-image.toml b/providers/aimlapi/models/alibaba/qwen3.6-max-preview.toml similarity index 54% rename from providers/aimlapi/models/google/gemini-3.1-flash-image.toml rename to providers/aimlapi/models/alibaba/qwen3.6-max-preview.toml index 6856e087191..e0067ed011f 100644 --- a/providers/aimlapi/models/google/gemini-3.1-flash-image.toml +++ b/providers/aimlapi/models/alibaba/qwen3.6-max-preview.toml @@ -1,9 +1,9 @@ -base_model = "google/gemini-3.1-flash-image" +base_model = "alibaba/qwen3.6-max-preview" [[reasoning_options]] type = "effort" values = ["none", "low", "medium", "high"] [cost] -input = 0.6877 -output = 4.1262 +input = 0.65 +output = 3.9 diff --git a/providers/aimlapi/models/alibaba/qwen3.8-2.4t-a95b.toml b/providers/aimlapi/models/alibaba/qwen3.8-2.4t-a95b.toml index 3ae6076152a..b7c4dffe97a 100644 --- a/providers/aimlapi/models/alibaba/qwen3.8-2.4t-a95b.toml +++ b/providers/aimlapi/models/alibaba/qwen3.8-2.4t-a95b.toml @@ -2,7 +2,7 @@ base_model = "alibaba/qwen3.8-2.4t-a95b" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["low", "medium", "high"] [cost] input = 3.4385 diff --git a/providers/aimlapi/models/arcee-ai/trinity-large-thinking.toml b/providers/aimlapi/models/arcee-ai/trinity-large-thinking.toml index 5d0980ba3eb..33031f2fb01 100644 --- a/providers/aimlapi/models/arcee-ai/trinity-large-thinking.toml +++ b/providers/aimlapi/models/arcee-ai/trinity-large-thinking.toml @@ -2,7 +2,7 @@ base_model = "arcee-ai/trinity-large-thinking" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["low", "medium", "high"] [cost] input = 0.302588 diff --git a/providers/aimlapi/models/google/gemini-2.5-flash-image.toml b/providers/aimlapi/models/google/gemini-2.5-flash-image.toml deleted file mode 100644 index cd9fc31bd11..00000000000 --- a/providers/aimlapi/models/google/gemini-2.5-flash-image.toml +++ /dev/null @@ -1,10 +0,0 @@ -base_model = "google/gemini-2.5-flash-image" - -[[reasoning_options]] -type = "effort" -values = ["none", "low", "medium", "high"] - -[cost] -input = 0.742716 -output = 6.1893 -cache_read = 0.0742716 diff --git a/providers/aimlapi/models/google/gemini-2.5-flash-lite.toml b/providers/aimlapi/models/google/gemini-2.5-flash-lite.toml index 84c6d6a63de..6675f8837ce 100644 --- a/providers/aimlapi/models/google/gemini-2.5-flash-lite.toml +++ b/providers/aimlapi/models/google/gemini-2.5-flash-lite.toml @@ -2,7 +2,7 @@ base_model = "google/gemini-2.5-flash-lite" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["minimal", "low", "medium", "high", "max"] [cost] input = 0.0975 @@ -11,4 +11,3 @@ cache_read = 0.0975 [limit] context = 1_000_000 -output = 1_048_576 diff --git a/providers/aimlapi/models/google/gemini-3-pro-image-preview.toml b/providers/aimlapi/models/google/gemini-3-pro-image-preview.toml deleted file mode 100644 index 4932eb4ba0e..00000000000 --- a/providers/aimlapi/models/google/gemini-3-pro-image-preview.toml +++ /dev/null @@ -1,10 +0,0 @@ -base_model = "google/gemini-3-pro-image-preview" - -[[reasoning_options]] -type = "effort" -values = ["none", "low", "medium", "high"] - -[cost] -input = 2.7508 -output = 16.5048 -cache_read = 0.27508 diff --git a/providers/aimlapi/models/google/gemini-3-pro-image.toml b/providers/aimlapi/models/google/gemini-3-pro-image.toml deleted file mode 100644 index 7f6c33369c9..00000000000 --- a/providers/aimlapi/models/google/gemini-3-pro-image.toml +++ /dev/null @@ -1,10 +0,0 @@ -base_model = "google/gemini-3-pro-image" - -[[reasoning_options]] -type = "effort" -values = ["none", "low", "medium", "high"] - -[cost] -input = 2.7508 -output = 16.5048 -cache_read = 0.27508 diff --git a/providers/aimlapi/models/google/gemini-3.1-flash-image-preview.toml b/providers/aimlapi/models/google/gemini-3.1-flash-image-preview.toml deleted file mode 100644 index 19d5157aacd..00000000000 --- a/providers/aimlapi/models/google/gemini-3.1-flash-image-preview.toml +++ /dev/null @@ -1,13 +0,0 @@ -base_model = "google/gemini-3.1-flash-image-preview" - -[[reasoning_options]] -type = "effort" -values = ["none", "low", "medium", "high"] - -[cost] -input = 0.6877 -output = 4.1262 - -[limit] -context = 131_072 -output = 32_768 diff --git a/providers/aimlapi/models/google/gemini-3.1-flash-lite-image.toml b/providers/aimlapi/models/google/gemini-3.1-flash-lite-image.toml deleted file mode 100644 index c9a308a289e..00000000000 --- a/providers/aimlapi/models/google/gemini-3.1-flash-lite-image.toml +++ /dev/null @@ -1,12 +0,0 @@ -base_model = "google/gemini-3.1-flash-lite-image" - -[[reasoning_options]] -type = "effort" -values = ["none", "low", "medium", "high"] - -[cost] -input = 0.34385 -output = 2.0631 - -[limit] -output = 66_000 diff --git a/providers/aimlapi/models/google/gemini-3.6-flash.toml b/providers/aimlapi/models/google/gemini-3.6-flash.toml index 8cce4b4e5db..25ed6371a38 100644 --- a/providers/aimlapi/models/google/gemini-3.6-flash.toml +++ b/providers/aimlapi/models/google/gemini-3.6-flash.toml @@ -2,7 +2,7 @@ base_model = "google/gemini-3.6-flash" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["minimal", "low", "medium", "high", "max"] [cost] input = 1.95 diff --git a/providers/aimlapi/models/google/gemini-3.7-flash.toml b/providers/aimlapi/models/google/gemini-3.7-flash.toml index c159244eee9..282c1f3d4bc 100644 --- a/providers/aimlapi/models/google/gemini-3.7-flash.toml +++ b/providers/aimlapi/models/google/gemini-3.7-flash.toml @@ -2,7 +2,7 @@ base_model = "google/gemini-3.7-flash" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["minimal", "low", "medium", "high", "max"] [cost] input = 0.975 diff --git a/providers/aimlapi/models/google/gemma-4-31b-it.toml b/providers/aimlapi/models/google/gemma-4-31b-it.toml index 19755a0cfa3..ed5edce7b43 100644 --- a/providers/aimlapi/models/google/gemma-4-31b-it.toml +++ b/providers/aimlapi/models/google/gemma-4-31b-it.toml @@ -11,4 +11,4 @@ cache_read = 1.361646 [limit] context = 256_000 -output = 262_144 +output = 16_384 diff --git a/providers/aimlapi/models/mistralai/mistral-nemo.toml b/providers/aimlapi/models/mistralai/mistral-nemo.toml index f762b3836f3..e635d20050c 100644 --- a/providers/aimlapi/models/mistralai/mistral-nemo.toml +++ b/providers/aimlapi/models/mistralai/mistral-nemo.toml @@ -6,4 +6,4 @@ output = 0.233818 cache_read = 0.20631 [limit] -output = 131_072 +output = 16_384 diff --git a/providers/aimlapi/models/moonshotai/kimi-k2-thinking.toml b/providers/aimlapi/models/moonshotai/kimi-k2-thinking.toml index ff971baa25e..30683df0f19 100644 --- a/providers/aimlapi/models/moonshotai/kimi-k2-thinking.toml +++ b/providers/aimlapi/models/moonshotai/kimi-k2-thinking.toml @@ -2,7 +2,7 @@ base_model = "moonshotai/kimi-k2-thinking" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["low", "medium", "high"] [cost] input = 0.82524 @@ -10,4 +10,4 @@ output = 3.4385 cache_read = 0.20631 [limit] -output = 100_352 +output = 98_304 diff --git a/providers/aimlapi/models/openai/gpt-5-mini.toml b/providers/aimlapi/models/openai/gpt-5-mini.toml index 69fae50bb44..112c96dff2a 100644 --- a/providers/aimlapi/models/openai/gpt-5-mini.toml +++ b/providers/aimlapi/models/openai/gpt-5-mini.toml @@ -2,7 +2,7 @@ base_model = "openai/gpt-5-mini" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["low", "medium", "high"] [cost] input = 0.378235 diff --git a/providers/aimlapi/models/openai/gpt-5-nano.toml b/providers/aimlapi/models/openai/gpt-5-nano.toml index 56bebab6927..8a081097106 100644 --- a/providers/aimlapi/models/openai/gpt-5-nano.toml +++ b/providers/aimlapi/models/openai/gpt-5-nano.toml @@ -2,7 +2,7 @@ base_model = "openai/gpt-5-nano" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["low", "medium", "high"] [cost] input = 0.075647 diff --git a/providers/aimlapi/models/openai/gpt-5-pro.toml b/providers/aimlapi/models/openai/gpt-5-pro.toml index ee8cb7ee067..dcf302b6018 100644 --- a/providers/aimlapi/models/openai/gpt-5-pro.toml +++ b/providers/aimlapi/models/openai/gpt-5-pro.toml @@ -2,7 +2,7 @@ base_model = "openai/gpt-5-pro" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["low", "medium", "high"] [cost] input = 20.631 diff --git a/providers/aimlapi/models/openai/gpt-5.1-codex.toml b/providers/aimlapi/models/openai/gpt-5.1-codex.toml index a46ce9b3fde..6c5c7db6021 100644 --- a/providers/aimlapi/models/openai/gpt-5.1-codex.toml +++ b/providers/aimlapi/models/openai/gpt-5.1-codex.toml @@ -2,7 +2,7 @@ base_model = "openai/gpt-5.1-codex" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["low", "medium", "high"] [cost] input = 3.4385 diff --git a/providers/aimlapi/models/openai/gpt-5.2-codex.toml b/providers/aimlapi/models/openai/gpt-5.2-codex.toml index 90bdfd92974..c84a5f445e5 100644 --- a/providers/aimlapi/models/openai/gpt-5.2-codex.toml +++ b/providers/aimlapi/models/openai/gpt-5.2-codex.toml @@ -2,7 +2,7 @@ base_model = "openai/gpt-5.2-codex" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["low", "medium", "high"] [cost] input = 4.8139 diff --git a/providers/aimlapi/models/openai/gpt-5.2-pro.toml b/providers/aimlapi/models/openai/gpt-5.2-pro.toml index bf69a62b5f1..d3e1d34fc0f 100644 --- a/providers/aimlapi/models/openai/gpt-5.2-pro.toml +++ b/providers/aimlapi/models/openai/gpt-5.2-pro.toml @@ -2,7 +2,7 @@ base_model = "openai/gpt-5.2-pro" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["low", "medium", "high"] [cost] input = 28.8834 diff --git a/providers/aimlapi/models/openai/gpt-5.4-pro.toml b/providers/aimlapi/models/openai/gpt-5.4-pro.toml index 8995a0fc847..07fcde9d4f9 100644 --- a/providers/aimlapi/models/openai/gpt-5.4-pro.toml +++ b/providers/aimlapi/models/openai/gpt-5.4-pro.toml @@ -2,7 +2,7 @@ base_model = "openai/gpt-5.4-pro" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["low", "medium", "high"] [cost] input = 41.262 diff --git a/providers/aimlapi/models/openai/gpt-5.5-pro.toml b/providers/aimlapi/models/openai/gpt-5.5-pro.toml index 3e419c38335..2002aaca7ae 100644 --- a/providers/aimlapi/models/openai/gpt-5.5-pro.toml +++ b/providers/aimlapi/models/openai/gpt-5.5-pro.toml @@ -2,7 +2,7 @@ base_model = "openai/gpt-5.5-pro" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["low", "medium", "high"] [cost] input = 41.262 diff --git a/providers/aimlapi/models/openai/gpt-5.toml b/providers/aimlapi/models/openai/gpt-5.toml index 71b1e3b3138..d202dbb58fc 100644 --- a/providers/aimlapi/models/openai/gpt-5.toml +++ b/providers/aimlapi/models/openai/gpt-5.toml @@ -2,7 +2,7 @@ base_model = "openai/gpt-5" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["low", "medium", "high"] [cost] input = 1.891175 diff --git a/providers/aimlapi/models/openai/gpt-oss-120b.toml b/providers/aimlapi/models/openai/gpt-oss-120b.toml index 55a8f241e13..312f08e3d34 100644 --- a/providers/aimlapi/models/openai/gpt-oss-120b.toml +++ b/providers/aimlapi/models/openai/gpt-oss-120b.toml @@ -2,7 +2,7 @@ base_model = "openai/gpt-oss-120b" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["low", "medium", "high"] [cost] input = 0.48139 @@ -11,4 +11,4 @@ cache_read = 0.48139 [limit] context = 131_000 -output = 131_072 +output = 117_964 diff --git a/providers/aimlapi/models/openai/gpt-oss-20b.toml b/providers/aimlapi/models/openai/gpt-oss-20b.toml index 68998c7a33e..c1755a91dbd 100644 --- a/providers/aimlapi/models/openai/gpt-oss-20b.toml +++ b/providers/aimlapi/models/openai/gpt-oss-20b.toml @@ -2,7 +2,7 @@ base_model = "openai/gpt-oss-20b" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["low", "medium", "high"] [cost] input = 0.103155 @@ -11,4 +11,4 @@ cache_read = 0.0515775 [limit] context = 131_000 -output = 131_072 +output = 117_964 diff --git a/providers/aimlapi/models/openai/o1-pro.toml b/providers/aimlapi/models/openai/o1-pro.toml index aee7f3e2ce5..11ce7cea6a2 100644 --- a/providers/aimlapi/models/openai/o1-pro.toml +++ b/providers/aimlapi/models/openai/o1-pro.toml @@ -2,7 +2,7 @@ base_model = "openai/o1-pro" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["low", "medium", "high"] [cost] input = 206.31 diff --git a/providers/aimlapi/models/openai/o1.toml b/providers/aimlapi/models/openai/o1.toml index 7704addd34a..95071e55c77 100644 --- a/providers/aimlapi/models/openai/o1.toml +++ b/providers/aimlapi/models/openai/o1.toml @@ -2,7 +2,7 @@ base_model = "openai/o1" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["low", "medium", "high", "xhigh"] [cost] input = 19.5 diff --git a/providers/aimlapi/models/openai/o3-mini.toml b/providers/aimlapi/models/openai/o3-mini.toml index 2f7f1df6a8a..469be08998e 100644 --- a/providers/aimlapi/models/openai/o3-mini.toml +++ b/providers/aimlapi/models/openai/o3-mini.toml @@ -2,7 +2,7 @@ base_model = "openai/o3-mini" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["low", "medium", "high", "xhigh"] [cost] input = 1.43 diff --git a/providers/aimlapi/models/openai/o3-pro.toml b/providers/aimlapi/models/openai/o3-pro.toml index 20e80f25699..7a5a8a0bd15 100644 --- a/providers/aimlapi/models/openai/o3-pro.toml +++ b/providers/aimlapi/models/openai/o3-pro.toml @@ -2,7 +2,7 @@ base_model = "openai/o3-pro" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["low", "medium", "high"] [cost] input = 27.508 diff --git a/providers/aimlapi/models/z-ai/glm-4.5v.toml b/providers/aimlapi/models/z-ai/glm-4.5v.toml index c745623a2ad..65d6f914112 100644 --- a/providers/aimlapi/models/z-ai/glm-4.5v.toml +++ b/providers/aimlapi/models/z-ai/glm-4.5v.toml @@ -2,7 +2,7 @@ base_model = "zhipuai/glm-4.5v" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["none", "minimal", "low", "medium", "high"] [cost] input = 0.82524 diff --git a/providers/aimlapi/models/z-ai/glm-4.6v.toml b/providers/aimlapi/models/z-ai/glm-4.6v.toml index ba8f70f7db0..953152d5a67 100644 --- a/providers/aimlapi/models/z-ai/glm-4.6v.toml +++ b/providers/aimlapi/models/z-ai/glm-4.6v.toml @@ -2,7 +2,7 @@ base_model = "zhipuai/glm-4.6v" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["none", "minimal", "low", "medium", "high"] [cost] input = 0.41262 diff --git a/providers/aimlapi/models/z-ai/glm-4.7-flash.toml b/providers/aimlapi/models/z-ai/glm-4.7-flash.toml index d828cf1badf..82aaef79286 100644 --- a/providers/aimlapi/models/z-ai/glm-4.7-flash.toml +++ b/providers/aimlapi/models/z-ai/glm-4.7-flash.toml @@ -2,7 +2,7 @@ base_model = "zhipuai/glm-4.7-flash" [[reasoning_options]] type = "effort" -values = ["none", "low", "medium", "high"] +values = ["none", "minimal", "low", "medium", "high"] [cost] input = 0.171925 diff --git a/providers/aimlapi/models/z-ai/glm-5.3-flash.toml b/providers/aimlapi/models/z-ai/glm-5.3-flash.toml new file mode 100644 index 00000000000..35f84724a33 --- /dev/null +++ b/providers/aimlapi/models/z-ai/glm-5.3-flash.toml @@ -0,0 +1,13 @@ +base_model = "zhipuai/glm-5.3-flash" + +[[reasoning_options]] +type = "effort" +values = ["low", "high", "max"] + +[cost] +input = 0.0975 +output = 0.325 +cache_read = 0.0195 + +[limit] +context = 1_048_576