diff --git a/providers/tinfoil/models/glm-5-3.toml b/providers/tinfoil/models/glm-5-3.toml new file mode 100644 index 00000000000..5452c715c46 --- /dev/null +++ b/providers/tinfoil/models/glm-5-3.toml @@ -0,0 +1,18 @@ +# Sources (accessed 2026-09-03): +# - https://inference.tinfoil.sh/v1/models (id glm-5-3; context 1048576; pricing input 1.8 / output 5.75 / cache_read 0.45 per 1M) +# - https://docs.tinfoil.sh/models/chat.md (GLM-5.3; 1M context; weights Inferact/GLM-5.3-NVFP4) +# - https://docs.tinfoil.sh/guides/reasoning.md (glm-5-3 effort low|high|max; reasoning always on) +# Effort: reasoning_effort = low|high|max on Tinfoil; reasoning is always on (no none; enable_thinking not honoured). +base_model = "zhipuai/glm-5.3" + +[[reasoning_options]] +type = "effort" +values = ["low", "high", "max"] + +[cost] +input = 1.8 +output = 5.75 +cache_read = 0.45 + +[limit] +context = 1_048_576