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<any>;
   ambient: SyncProvider<any>;
   anthropic: SyncProvider<any>;
   baseten: SyncProvider<any>;
@@ -162,6 +164,7 @@ export const providers: {
   wandb: SyncProvider<any>;
   xai: SyncProvider<any>;
 } = {
+  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..d8c7b25c185
--- /dev/null
+++ b/packages/core/src/sync/providers/aimlapi.ts
@@ -0,0 +1,330 @@
+import { z } from "zod";
+
+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";
+
+// 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<typeof AimlapiModel>;
+
+type Modality = "text" | "audio" | "image" | "video" | "pdf";
+
+const MODALITIES = new Set<string>(["text", "audio", "image", "video", "pdf"]);
+
+function normalizeModalities(values: readonly string[] | null | undefined): Modality[] {
+  const seen = new Set<Modality>();
+  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];
+}
+
+/**
+ * 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<string>();
+
+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.
+  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<typeof PricingUnit>[], 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;
+}
+
+/**
+ * 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".
+ */
+async function fetchReasoningEffort(id: string): Promise<string[] | undefined> {
+  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 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<string, unknown>;
+  const effort = record["reasoning_effort"];
+  if (effort !== null && typeof effort === "object") {
+    const values = (effort as Record<string, unknown>)["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<void> {
+  // 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 = {
+  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 isChatTextModel(model) ? model.id : undefined;
+  },
+  skippedNotice(ids) {
+    if (ids.length === 0) return [];
+    return [
+      `${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(", ")}`,
+    ];
+  },
+  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}`);
+    }
+    const raw = await response.json();
+    const parsed = AimlapiResponse.parse(raw);
+    indexMediaOutputs(parsed.data);
+    await attachReasoningEffort(parsed.data);
+    return parsed;
+  },
+  parseModels(raw) {
+    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;
+
+    const existing = context.existing(model.id);
+
+    // 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;
+
+    // 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 }];
+    }
+
+    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: 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<AimlapiModel>;
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 @@
+<svg viewBox="0 0 16 16" fill="none" xmlns="http://www.w3.org/2000/svg">
+  <path fill-rule="evenodd" clip-rule="evenodd" d="M 15.2039 7.1382 C 15.4949 7.6422 15.4949 8.2632 15.2039 8.7672 L 12.1861 13.9943 C 11.8951 14.4983 11.3573 14.8089 10.7753 14.8089 H 4.7396 C 4.1576 14.8089 3.6198 14.4983 3.3288 13.9943 L 0.3109 8.7672 C 0.0199 8.2632 0.0199 7.6422 0.3109 7.1382 L 3.3288 1.9111 C 3.6198 1.4071 4.1576 1.0966 4.7396 1.0966 H 10.7753 C 11.3573 1.0966 11.8951 1.4071 12.1861 1.9111 L 15.2039 7.1382 Z M 4.4453 10.2545 C 4.3318 10.2545 4.2176 10.2183 4.1211 10.1445 C 3.8884 9.9651 3.8444 9.6309 4.0239 9.3983 L 6.1471 6.6387 C 6.2337 6.5259 6.3621 6.4529 6.5025 6.4351 C 6.6458 6.4167 6.7863 6.4564 6.8977 6.5451 L 8.8981 8.1164 L 10.6483 5.8584 C 10.8285 5.625 11.1619 5.5817 11.3945 5.7633 C 11.6272 5.9435 11.6698 6.2776 11.4896 6.5096 L 9.4111 9.1911 C 9.3245 9.3032 9.1968 9.3763 9.0563 9.3933 C 8.9145 9.4125 8.774 9.3713 8.6619 9.2841 L 6.6629 7.7135 L 4.8674 10.0467 C 4.7624 10.1829 4.6049 10.2545 4.4453 10.2545 Z" fill="currentColor"/>
+</svg>
diff --git a/providers/aimlapi/models/alibaba/qwen-max.toml b/providers/aimlapi/models/alibaba/qwen-max.toml
new file mode 100644
index 00000000000..b67b2dfda47
--- /dev/null
+++ b/providers/aimlapi/models/alibaba/qwen-max.toml
@@ -0,0 +1,9 @@
+base_model = "alibaba/qwen-max"
+
+[cost]
+input = 2.08
+output = 8.32
+cache_read = 0.416
+
+[limit]
+context = 32_000
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..c294f7401f9
--- /dev/null
+++ b/providers/aimlapi/models/alibaba/qwen3-coder-480b-a35b-instruct.toml
@@ -0,0 +1,8 @@
+base_model = "alibaba/qwen3-coder-480b-a35b-instruct"
+
+[cost]
+input = 1.95
+output = 9.75
+
+[limit]
+context = 262_000
diff --git a/providers/aimlapi/models/alibaba/qwen3-max.toml b/providers/aimlapi/models/alibaba/qwen3-max.toml
new file mode 100644
index 00000000000..2cc25e4c73e
--- /dev/null
+++ b/providers/aimlapi/models/alibaba/qwen3-max.toml
@@ -0,0 +1,6 @@
+base_model = "alibaba/qwen3-max"
+
+[cost]
+input = 1.56
+output = 7.8
+cache_read = 0.312
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..b90216744f6
--- /dev/null
+++ b/providers/aimlapi/models/alibaba/qwen3-next-80b-a3b-instruct.toml
@@ -0,0 +1,9 @@
+base_model = "alibaba/qwen3-next-80b-a3b-instruct"
+
+[cost]
+input = 0.195
+output = 1.56
+
+[limit]
+context = 129_024
+output = 16_384
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..ce03c2805d0
--- /dev/null
+++ b/providers/aimlapi/models/alibaba/qwen3.5-flash.toml
@@ -0,0 +1,9 @@
+base_model = "alibaba/qwen3.5-flash"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high"]
+
+[cost]
+input = 0.13
+output = 0.52
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-max-preview.toml b/providers/aimlapi/models/alibaba/qwen3.6-max-preview.toml
new file mode 100644
index 00000000000..e0067ed011f
--- /dev/null
+++ b/providers/aimlapi/models/alibaba/qwen3.6-max-preview.toml
@@ -0,0 +1,9 @@
+base_model = "alibaba/qwen3.6-max-preview"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high"]
+
+[cost]
+input = 0.65
+output = 3.9
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..ad69c73c11c
--- /dev/null
+++ b/providers/aimlapi/models/alibaba/qwen3.7-max.toml
@@ -0,0 +1,10 @@
+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
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..b7c4dffe97a
--- /dev/null
+++ b/providers/aimlapi/models/alibaba/qwen3.8-2.4t-a95b.toml
@@ -0,0 +1,13 @@
+base_model = "alibaba/qwen3.8-2.4t-a95b"
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high"]
+
+[cost]
+input = 3.4385
+output = 8.59625
+cache_read = 0.6877
+
+[limit]
+output = 52_429
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..77fdf0eda88
--- /dev/null
+++ b/providers/aimlapi/models/alibaba/qwen3.8-27b.toml
@@ -0,0 +1,12 @@
+base_model = "alibaba/qwen3.8-27b"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high"]
+
+[cost]
+input = 0.61893
+output = 4.40128
+
+[limit]
+output = 131_072
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..811f1426cff
--- /dev/null
+++ b/providers/aimlapi/models/alibaba/qwen3.8-flash.toml
@@ -0,0 +1,10 @@
+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
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..654d7093d5e
--- /dev/null
+++ b/providers/aimlapi/models/alibaba/qwen3.8-max.toml
@@ -0,0 +1,10 @@
+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
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..626bc3d762f
--- /dev/null
+++ b/providers/aimlapi/models/anthropic/claude-haiku-4.5.toml
@@ -0,0 +1,10 @@
+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
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..c86491a03be
--- /dev/null
+++ b/providers/aimlapi/models/anthropic/claude-opus-4.1.toml
@@ -0,0 +1,10 @@
+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
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..1d5bfafae16
--- /dev/null
+++ b/providers/aimlapi/models/anthropic/claude-opus-4.5.toml
@@ -0,0 +1,10 @@
+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
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
new file mode 100644
index 00000000000..f57da60fd4f
--- /dev/null
+++ b/providers/aimlapi/models/anthropic/claude-opus-4.8.toml
@@ -0,0 +1,10 @@
+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
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..abdf9b07883
--- /dev/null
+++ b/providers/aimlapi/models/anthropic/claude-sonnet-4.5.toml
@@ -0,0 +1,10 @@
+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
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..a46dbba1d12
--- /dev/null
+++ b/providers/aimlapi/models/anthropic/claude-sonnet-4.6.toml
@@ -0,0 +1,13 @@
+base_model = "anthropic/claude-sonnet-4-6"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high"]
+
+[cost]
+input = 4.1262
+output = 20.631
+cache_read = 0.41262
+
+[limit]
+context = 200_000
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..33031f2fb01
--- /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 = ["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/bytedance-seed/seed-2.0-code.toml b/providers/aimlapi/models/bytedance-seed/seed-2.0-code.toml
new file mode 100644
index 00000000000..b22d70d2be7
--- /dev/null
+++ b/providers/aimlapi/models/bytedance-seed/seed-2.0-code.toml
@@ -0,0 +1,9 @@
+base_model = "bytedance-seed/seed-2.0-code"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high"]
+
+[cost]
+input = 0.6877
+output = 4.1262
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/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.toml b/providers/aimlapi/models/deepseek/deepseek-chat.toml
new file mode 100644
index 00000000000..f47b52b0ab1
--- /dev/null
+++ b/providers/aimlapi/models/deepseek/deepseek-chat.toml
@@ -0,0 +1,10 @@
+base_model = "deepseek/deepseek-chat"
+
+[cost]
+input = 0.364
+output = 0.546
+cache_read = 0.0364
+
+[limit]
+context = 128_000
+output = 124_000
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..14c8b06da73
--- /dev/null
+++ b/providers/aimlapi/models/deepseek/deepseek-v4-flash-vision-exp.toml
@@ -0,0 +1,13 @@
+base_model = "deepseek/deepseek-v4-flash-vision-exp"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high"]
+
+[cost]
+input = 0.572
+output = 1.716
+cache_read = 0.0182
+
+[limit]
+context = 1_048_576
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..56437429724
--- /dev/null
+++ b/providers/aimlapi/models/deepseek/deepseek-v4-flash.toml
@@ -0,0 +1,10 @@
+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
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..c8947cb45c0
--- /dev/null
+++ b/providers/aimlapi/models/deepseek/deepseek-v4-pro-0813.toml
@@ -0,0 +1,13 @@
+base_model = "deepseek/deepseek-v4-pro-0813"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high"]
+
+[cost]
+input = 2.393196
+output = 4.786392
+cache_read = 0.199433
+
+[limit]
+context = 1_048_576
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..e733d802ed8
--- /dev/null
+++ b/providers/aimlapi/models/deepseek/deepseek-v4-pro.toml
@@ -0,0 +1,10 @@
+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
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..6675f8837ce
--- /dev/null
+++ b/providers/aimlapi/models/google/gemini-2.5-flash-lite.toml
@@ -0,0 +1,13 @@
+base_model = "google/gemini-2.5-flash-lite"
+
+[[reasoning_options]]
+type = "effort"
+values = ["minimal", "low", "medium", "high", "max"]
+
+[cost]
+input = 0.0975
+output = 0.39
+cache_read = 0.0975
+
+[limit]
+context = 1_000_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-3.6-flash.toml b/providers/aimlapi/models/google/gemini-3.6-flash.toml
new file mode 100644
index 00000000000..25ed6371a38
--- /dev/null
+++ b/providers/aimlapi/models/google/gemini-3.6-flash.toml
@@ -0,0 +1,10 @@
+base_model = "google/gemini-3.6-flash"
+
+[[reasoning_options]]
+type = "effort"
+values = ["minimal", "low", "medium", "high", "max"]
+
+[cost]
+input = 1.95
+output = 9.75
+cache_read = 0.195
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..282c1f3d4bc
--- /dev/null
+++ b/providers/aimlapi/models/google/gemini-3.7-flash.toml
@@ -0,0 +1,10 @@
+base_model = "google/gemini-3.7-flash"
+
+[[reasoning_options]]
+type = "effort"
+values = ["minimal", "low", "medium", "high", "max"]
+
+[cost]
+input = 0.975
+output = 4.875
+cache_read = 0.0975
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-4-26b-a4b-it.toml b/providers/aimlapi/models/google/gemma-4-26b-a4b-it.toml
new file mode 100644
index 00000000000..86370326cb4
--- /dev/null
+++ b/providers/aimlapi/models/google/gemma-4-26b-a4b-it.toml
@@ -0,0 +1,13 @@
+base_model = "google/gemma-4-26b-a4b-it"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high"]
+
+[cost]
+input = 0.20631
+output = 0.82524
+cache_read = 0.20631
+
+[limit]
+output = 128_000
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..ed5edce7b43
--- /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 = 16_384
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/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
new file mode 100644
index 00000000000..e635d20050c
--- /dev/null
+++ b/providers/aimlapi/models/mistralai/mistral-nemo.toml
@@ -0,0 +1,9 @@
+base_model = "mistral/mistral-nemo"
+
+[cost]
+input = 0.20631
+output = 0.233818
+cache_read = 0.20631
+
+[limit]
+output = 16_384
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/moonshotai/kimi-k2-thinking.toml b/providers/aimlapi/models/moonshotai/kimi-k2-thinking.toml
new file mode 100644
index 00000000000..30683df0f19
--- /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 = ["low", "medium", "high"]
+
+[cost]
+input = 0.82524
+output = 3.4385
+cache_read = 0.20631
+
+[limit]
+output = 98_304
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-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
new file mode 100644
index 00000000000..e522294beb7
--- /dev/null
+++ b/providers/aimlapi/models/openai/gpt-4.1.toml
@@ -0,0 +1,6 @@
+base_model = "openai/gpt-4.1"
+
+[cost]
+input = 2.6
+output = 10.4
+cache_read = 0.65
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
new file mode 100644
index 00000000000..fbfb5174ab3
--- /dev/null
+++ b/providers/aimlapi/models/openai/gpt-4o-2024-05-13.toml
@@ -0,0 +1,8 @@
+base_model = "openai/gpt-4o-2024-05-13"
+
+[cost]
+input = 6.5
+output = 19.5
+
+[limit]
+output = 16_384
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..37cdb080980
--- /dev/null
+++ b/providers/aimlapi/models/openai/gpt-4o-2024-08-06.toml
@@ -0,0 +1,6 @@
+base_model = "openai/gpt-4o-2024-08-06"
+
+[cost]
+input = 3.25
+output = 13
+cache_read = 1.625
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..909b9284067
--- /dev/null
+++ b/providers/aimlapi/models/openai/gpt-4o-2024-11-20.toml
@@ -0,0 +1,6 @@
+base_model = "openai/gpt-4o-2024-11-20"
+
+[cost]
+input = 3.25
+output = 13
+cache_read = 1.625
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
new file mode 100644
index 00000000000..88760223978
--- /dev/null
+++ b/providers/aimlapi/models/openai/gpt-4o.toml
@@ -0,0 +1,6 @@
+base_model = "openai/gpt-4o"
+
+[cost]
+input = 3.25
+output = 13
+cache_read = 1.625
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..112c96dff2a
--- /dev/null
+++ b/providers/aimlapi/models/openai/gpt-5-mini.toml
@@ -0,0 +1,10 @@
+base_model = "openai/gpt-5-mini"
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high"]
+
+[cost]
+input = 0.378235
+output = 3.02588
+cache_read = 0.34385
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..8a081097106
--- /dev/null
+++ b/providers/aimlapi/models/openai/gpt-5-nano.toml
@@ -0,0 +1,10 @@
+base_model = "openai/gpt-5-nano"
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high"]
+
+[cost]
+input = 0.075647
+output = 0.605176
+cache_read = 0.06877
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..dcf302b6018
--- /dev/null
+++ b/providers/aimlapi/models/openai/gpt-5-pro.toml
@@ -0,0 +1,9 @@
+base_model = "openai/gpt-5-pro"
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high"]
+
+[cost]
+input = 20.631
+output = 165.048
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
new file mode 100644
index 00000000000..fa3cd2062d2
--- /dev/null
+++ b/providers/aimlapi/models/openai/gpt-5.1-codex-mini.toml
@@ -0,0 +1,10 @@
+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
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..6c5c7db6021
--- /dev/null
+++ b/providers/aimlapi/models/openai/gpt-5.1-codex.toml
@@ -0,0 +1,10 @@
+base_model = "openai/gpt-5.1-codex"
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high"]
+
+[cost]
+input = 3.4385
+output = 27.508
+cache_read = 0.34385
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..c84a5f445e5
--- /dev/null
+++ b/providers/aimlapi/models/openai/gpt-5.2-codex.toml
@@ -0,0 +1,10 @@
+base_model = "openai/gpt-5.2-codex"
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high"]
+
+[cost]
+input = 4.8139
+output = 38.5112
+cache_read = 0.48139
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..d3e1d34fc0f
--- /dev/null
+++ b/providers/aimlapi/models/openai/gpt-5.2-pro.toml
@@ -0,0 +1,9 @@
+base_model = "openai/gpt-5.2-pro"
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high"]
+
+[cost]
+input = 28.8834
+output = 231.0672
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..f69e9a098bf
--- /dev/null
+++ b/providers/aimlapi/models/openai/gpt-5.3-codex.toml
@@ -0,0 +1,10 @@
+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
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..07fcde9d4f9
--- /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 = ["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..2002aaca7ae
--- /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 = ["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
new file mode 100644
index 00000000000..d202dbb58fc
--- /dev/null
+++ b/providers/aimlapi/models/openai/gpt-5.toml
@@ -0,0 +1,10 @@
+base_model = "openai/gpt-5"
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high"]
+
+[cost]
+input = 1.891175
+output = 15.1294
+cache_read = 1.71925
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..312f08e3d34
--- /dev/null
+++ b/providers/aimlapi/models/openai/gpt-oss-120b.toml
@@ -0,0 +1,14 @@
+base_model = "openai/gpt-oss-120b"
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high"]
+
+[cost]
+input = 0.48139
+output = 1.30663
+cache_read = 0.48139
+
+[limit]
+context = 131_000
+output = 117_964
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..c1755a91dbd
--- /dev/null
+++ b/providers/aimlapi/models/openai/gpt-oss-20b.toml
@@ -0,0 +1,14 @@
+base_model = "openai/gpt-oss-20b"
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high"]
+
+[cost]
+input = 0.103155
+output = 0.41262
+cache_read = 0.0515775
+
+[limit]
+context = 131_000
+output = 117_964
diff --git a/providers/aimlapi/models/openai/o1-pro.toml b/providers/aimlapi/models/openai/o1-pro.toml
new file mode 100644
index 00000000000..11ce7cea6a2
--- /dev/null
+++ b/providers/aimlapi/models/openai/o1-pro.toml
@@ -0,0 +1,9 @@
+base_model = "openai/o1-pro"
+
+[[reasoning_options]]
+type = "effort"
+values = ["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..95071e55c77
--- /dev/null
+++ b/providers/aimlapi/models/openai/o1.toml
@@ -0,0 +1,13 @@
+base_model = "openai/o1"
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high", "xhigh"]
+
+[cost]
+input = 19.5
+output = 78
+cache_read = 9.75
+
+[limit]
+context = 128_000
diff --git a/providers/aimlapi/models/openai/o3-mini.toml b/providers/aimlapi/models/openai/o3-mini.toml
new file mode 100644
index 00000000000..469be08998e
--- /dev/null
+++ b/providers/aimlapi/models/openai/o3-mini.toml
@@ -0,0 +1,10 @@
+base_model = "openai/o3-mini"
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high", "xhigh"]
+
+[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..7a5a8a0bd15
--- /dev/null
+++ b/providers/aimlapi/models/openai/o3-pro.toml
@@ -0,0 +1,9 @@
+base_model = "openai/o3-pro"
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high"]
+
+[cost]
+input = 27.508
+output = 110.032
diff --git a/providers/aimlapi/models/perplexity/sonar-pro.toml b/providers/aimlapi/models/perplexity/sonar-pro.toml
new file mode 100644
index 00000000000..d353019ac3a
--- /dev/null
+++ b/providers/aimlapi/models/perplexity/sonar-pro.toml
@@ -0,0 +1,8 @@
+base_model = "perplexity/sonar-pro"
+
+[cost]
+input = 3.9
+output = 19.5
+
+[limit]
+output = 100_000
diff --git a/providers/aimlapi/models/perplexity/sonar.toml b/providers/aimlapi/models/perplexity/sonar.toml
new file mode 100644
index 00000000000..8b062038318
--- /dev/null
+++ b/providers/aimlapi/models/perplexity/sonar.toml
@@ -0,0 +1,8 @@
+base_model = "perplexity/sonar"
+
+[cost]
+input = 1.3
+output = 1.3
+
+[limit]
+output = 100_000
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
new file mode 100644
index 00000000000..f14cb9af343
--- /dev/null
+++ b/providers/aimlapi/models/sakana/fugu-ultra.toml
@@ -0,0 +1,13 @@
+base_model = "sakana/fugu-ultra"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high"]
+
+[cost]
+input = 6.877
+output = 41.262
+cache_read = 0.6877
+
+[limit]
+output = 128_000
diff --git a/providers/aimlapi/models/sakana/sakana-namazu.toml b/providers/aimlapi/models/sakana/sakana-namazu.toml
new file mode 100644
index 00000000000..0910a71e8c8
--- /dev/null
+++ b/providers/aimlapi/models/sakana/sakana-namazu.toml
@@ -0,0 +1,10 @@
+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
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
new file mode 100644
index 00000000000..cc65e8c53b8
--- /dev/null
+++ b/providers/aimlapi/models/tencent/hy4-preview.toml
@@ -0,0 +1,13 @@
+base_model = "tencent/hy4-preview"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high"]
+
+[cost]
+input = 1.1470836
+output = 3.4398754
+cache_read = 0.0577668
+
+[limit]
+context = 1_048_576
diff --git a/providers/aimlapi/models/thinkingmachines/inkling-small.toml b/providers/aimlapi/models/thinkingmachines/inkling-small.toml
new file mode 100644
index 00000000000..8ee54fbd811
--- /dev/null
+++ b/providers/aimlapi/models/thinkingmachines/inkling-small.toml
@@ -0,0 +1,14 @@
+base_model = "thinkingmachines/inkling-small"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high"]
+
+[cost]
+input = 0.797732
+output = 1.980576
+cache_read = 0.1595464
+
+[limit]
+context = 524_288
+output = 262_144
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
new file mode 100644
index 00000000000..ac67690de8d
--- /dev/null
+++ b/providers/aimlapi/models/upstage/solar-pro4.toml
@@ -0,0 +1,10 @@
+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
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..649a50bdb63
--- /dev/null
+++ b/providers/aimlapi/models/xiaomi/mimo-v2.5-pro.toml
@@ -0,0 +1,14 @@
+base_model = "xiaomi/mimo-v2.5-pro"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high"]
+
+[cost]
+input = 0.598299
+output = 1.196598
+cache_read = 0.00495144
+
+[limit]
+context = 1_000_000
+output = 128_000
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..8ea2de6fccc
--- /dev/null
+++ b/providers/aimlapi/models/xiaomi/mimo-v2.5.toml
@@ -0,0 +1,14 @@
+base_model = "xiaomi/mimo-v2.5"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high"]
+
+[cost]
+input = 0.192556
+output = 0.385112
+cache_read = 0.0038511200000000004
+
+[limit]
+context = 1_000_000
+output = 128_000
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..65d6f914112
--- /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", "minimal", "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..953152d5a67
--- /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", "minimal", "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..82aaef79286
--- /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", "minimal", "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
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
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"