diff --git a/packages/core/src/sync/index.ts b/packages/core/src/sync/index.ts
index 3b8c8fce97..72452d3e64 100644
--- a/packages/core/src/sync/index.ts
+++ b/packages/core/src/sync/index.ts
@@ -21,6 +21,7 @@ import { openai } from "./providers/openai.js";
 import { openrouter } from "./providers/openrouter.js";
 import { ovhcloud } from "./providers/ovhcloud.js";
 import { pioneer } from "./providers/pioneer.js";
+import { requesty } from "./providers/requesty.js";
 import { vercel } from "./providers/vercel.js";
 import { venice } from "./providers/venice.js";
 import { wandb } from "./providers/wandb.js";
@@ -110,6 +111,7 @@ export const providers: {
   openrouter: SyncProvider<any>;
   ovhcloud: SyncProvider<any>;
   pioneer: SyncProvider<any>;
+  requesty: SyncProvider<any>;
   vercel: SyncProvider<any>;
   venice: SyncProvider<any>;
   wandb: SyncProvider<any>;
@@ -132,6 +134,7 @@ export const providers: {
   openrouter,
   ovhcloud,
   pioneer,
+  requesty,
   vercel,
   venice,
   wandb,
@@ -139,7 +142,7 @@ export const providers: {
 };
 
 export const groups = {
-  aggregators: ["crossmodel", "empiriolabs", "huggingface", "kilo", "llmgateway", "openrouter", "vercel"],
+  aggregators: ["crossmodel", "empiriolabs", "huggingface", "kilo", "llmgateway", "openrouter", "requesty", "vercel"],
   cloudflare: ["cloudflare-workers-ai"],
   direct: ["ambient", "anthropic", "baseten", "chutes", "deepinfra", "digitalocean", "google", "openai", "ovhcloud", "pioneer", "venice", "wandb", "xai"],
 } as const;
diff --git a/packages/core/src/sync/providers/requesty.ts b/packages/core/src/sync/providers/requesty.ts
new file mode 100644
index 0000000000..3bf646d00b
--- /dev/null
+++ b/packages/core/src/sync/providers/requesty.ts
@@ -0,0 +1,259 @@
+import { readFileSync } from "node:fs";
+import path from "node:path";
+import { z } from "zod";
+
+import type {
+  ExistingModel,
+  SyncProvider,
+  SyncedFullModel,
+  SyncedModel,
+} from "../index.js";
+import { factorBaseModel, resolveCanonicalBaseModel } from "./openrouter.js";
+
+const API_ENDPOINT = "https://router.requesty.ai/v1/models";
+const MODELS_DIR = path.join(
+  import.meta.dirname,
+  "..",
+  "..",
+  "..",
+  "..",
+  "..",
+  "models",
+);
+const metadataCache = new Map<string, Record<string, unknown>>();
+const RESPONSES_MODELS = new Map([
+  ["openai-responses/gpt-5.4-mini", "openai/gpt-5.4-mini"],
+  ["openai-responses/gpt-5.4-nano", "openai/gpt-5.4-nano"],
+  ["openai-responses/gpt-5.5", "openai/gpt-5.5"],
+  ["openai-responses/gpt-5.6-luna", "openai/gpt-5.6-luna"],
+  ["openai-responses/gpt-5.6-sol", "openai/gpt-5.6-sol"],
+  ["openai-responses/gpt-5.6-terra", "openai/gpt-5.6-terra"],
+]);
+const SKIPPED_MODELS = new Set([
+  ...RESPONSES_MODELS.values(),
+  "nvidia/nemotron-3.5-content-safety",
+]);
+
+export const RequestyModel = z.object({
+  id: z.string(),
+  api: z.string().optional(),
+  object: z.string().optional(),
+  created: z.number().optional(),
+  description: z.string().optional(),
+  input_price: z.number().nonnegative().optional(),
+  output_price: z.number().nonnegative().optional(),
+  caching_price: z.number().nonnegative().optional(),
+  cached_price: z.number().nonnegative().optional(),
+  context_window: z.number().int().nonnegative().optional(),
+  max_output_tokens: z.number().int().nonnegative().optional(),
+  supports_caching: z.boolean().optional(),
+  supports_vision: z.boolean().optional(),
+  supports_reasoning: z.boolean().optional(),
+  supports_tool_calling: z.boolean().optional(),
+  supports_image_generation: z.boolean().optional(),
+  supports_output_json_object: z.boolean().optional(),
+  supports_output_json_schema: z.boolean().optional(),
+}).passthrough();
+
+export const RequestyResponse = z.object({
+  data: z.array(RequestyModel),
+}).passthrough();
+
+export type RequestyModel = z.infer<typeof RequestyModel>;
+
+export const requesty = {
+  id: "requesty",
+  name: "Requesty",
+  modelsDir: "providers/requesty/models",
+  // Retain curated aliases and models missing from the public catalog.
+  deleteMissing: false,
+  async fetchModels() {
+    const response = await fetch(API_ENDPOINT);
+    if (!response.ok) {
+      throw new Error(`Requesty request failed: ${response.status} ${response.statusText}`);
+    }
+    return response.json();
+  },
+  parseModels(raw) {
+    return RequestyResponse.parse(raw).data.filter((model) => {
+      if ((model.api ?? "chat") !== "chat") return false;
+      if (!model.id.includes("/") || model.id.startsWith("policy/")) return false;
+      if (SKIPPED_MODELS.has(model.id)) return false;
+      return requestyBaseModel(model.id) !== undefined;
+    });
+  },
+  translateModel(model, context) {
+    const existing = context.existing(model.id);
+    const authored = context.authored(model.id);
+    const canonical = authored?.base_model ?? requestyBaseModel(model.id);
+    if (canonical === undefined) return undefined;
+    return {
+      id: model.id,
+      model: buildRequestyModel(model, existing, authored, canonical),
+    };
+  },
+} satisfies SyncProvider<RequestyModel>;
+
+type Modality = "text" | "audio" | "image" | "video" | "pdf";
+
+function price(value: number | undefined) {
+  if (value === undefined) return undefined;
+  return Math.round(value * 1_000_000_000_000) / 1_000_000;
+}
+
+function metadata(modelID: string) {
+  let value = metadataCache.get(modelID);
+  if (value !== undefined) return value;
+  value = Bun.TOML.parse(
+    readFileSync(path.join(MODELS_DIR, `${modelID}.toml`), "utf8"),
+  ) as Record<string, unknown>;
+  metadataCache.set(modelID, value);
+  return value;
+}
+
+function metadataBoolean(modelID: string, field: string) {
+  const value = metadata(modelID)[field];
+  return typeof value === "boolean" ? value : undefined;
+}
+
+function metadataLimit(modelID: string) {
+  const value = metadata(modelID).limit;
+  return value !== null && typeof value === "object" && !Array.isArray(value)
+    ? value as { context?: number; input?: number; output?: number }
+    : undefined;
+}
+
+function metadataModalities(modelID: string) {
+  const value = metadata(modelID).modalities;
+  if (value === null || typeof value !== "object" || Array.isArray(value)) return undefined;
+  const modalities = value as { input?: Modality[]; output?: Modality[] };
+  return modalities.input !== undefined && modalities.output !== undefined
+    ? { input: modalities.input, output: modalities.output }
+    : undefined;
+}
+
+function append(values: Modality[], value: Modality, enabled: boolean | undefined) {
+  return enabled === true && !values.includes(value) ? [...values, value] : values;
+}
+
+function requestyBaseModel(modelID: string) {
+  return RESPONSES_MODELS.get(modelID) ?? resolveCanonicalBaseModel(modelID);
+}
+
+function reasoningOptions(
+  model: RequestyModel,
+  reasoning: boolean,
+  existing: ExistingModel | undefined,
+): SyncedFullModel["reasoning_options"] {
+  if (!reasoning) return undefined;
+  const canonical = requestyBaseModel(model.id);
+  const provider = canonical?.split("/")[0];
+  if (
+    provider === "openai"
+    || provider === "anthropic"
+    || canonical?.startsWith("google/gemini-")
+  ) {
+    return [
+      { type: "effort", values: ["none", "low", "medium", "high", "max"] },
+      { type: "budget_tokens" },
+    ];
+  }
+  if (canonical?.startsWith("google/gemma-")) return [];
+  return existing?.reasoning_options ?? [];
+}
+
+export function buildRequestyModel(
+  model: RequestyModel,
+  existing: ExistingModel | undefined,
+  authored: ExistingModel | undefined,
+  canonical = requestyBaseModel(model.id),
+): SyncedModel {
+  if (canonical === undefined) {
+    throw new Error(`Requesty model ${model.id} does not resolve to shared metadata`);
+  }
+
+  const baseLimit = metadataLimit(canonical);
+  const baseModalities = metadataModalities(canonical);
+  const inputModalities = [...new Set<Modality>([
+    ...(baseModalities?.input ?? []),
+    ...(existing?.modalities?.input ?? []),
+  ])];
+  const outputModalities = [...new Set<Modality>([
+    ...(baseModalities?.output ?? []),
+    ...(existing?.modalities?.output ?? []),
+  ])];
+  const modalities = {
+    input: append(inputModalities.length > 0 ? inputModalities : ["text"], "image", model.supports_vision),
+    output: append(outputModalities.length > 0 ? outputModalities : ["text"], "image", model.supports_image_generation),
+  };
+  const attachment = modalities.input.some((value) => value !== "text");
+  const intrinsicReasoning = existing?.reasoning === true
+    || metadataBoolean(canonical, "reasoning") === true;
+  const reasoning = intrinsicReasoning || model.supports_reasoning === true;
+  const toolCall = model.id === "google/gemini-3.1-flash-image-preview"
+    ? false
+    : model.supports_tool_calling
+      ?? existing?.tool_call
+      ?? metadataBoolean(canonical, "tool_call");
+  const reportedStructuredOutput = model.supports_output_json_schema === undefined
+      && model.supports_output_json_object === undefined
+    ? undefined
+    : model.supports_output_json_schema === true
+      || model.supports_output_json_object === true;
+  const structuredOutput = reportedStructuredOutput
+    ?? existing?.structured_output
+    ?? metadataBoolean(canonical, "structured_output");
+  const inputPrice = price(model.input_price);
+  const outputPrice = price(model.output_price);
+  const cost = inputPrice !== undefined && outputPrice !== undefined
+    ? {
+        input: inputPrice,
+        output: outputPrice,
+        cache_read: model.supports_caching === false
+          ? undefined
+          : price(model.cached_price) ?? existing?.cost?.cache_read,
+        cache_write: model.supports_caching === false
+          ? undefined
+          : price(model.caching_price) ?? existing?.cost?.cache_write,
+        tiers: existing?.cost?.tiers,
+      }
+    : existing?.cost;
+  const context = model.context_window ?? existing?.limit?.context ?? baseLimit?.context ?? 0;
+  const output = model.max_output_tokens !== undefined && model.max_output_tokens > 0
+    ? model.max_output_tokens
+    : existing?.limit?.output ?? baseLimit?.output ?? context;
+  const limit = {
+    context,
+    input: existing?.limit?.input ?? baseLimit?.input,
+    output,
+  };
+  const values: Partial<SyncedFullModel> = {
+    attachment,
+    reasoning,
+    reasoning_options: reasoningOptions(model, reasoning, existing),
+    temperature: existing?.temperature,
+    tool_call: toolCall,
+    structured_output: structuredOutput,
+    status: existing?.status,
+    interleaved: existing?.interleaved,
+    provider: RESPONSES_MODELS.has(model.id)
+      ? { npm: "@ai-sdk/openai", shape: "responses" }
+      : existing?.provider,
+    cost,
+    limit,
+    modalities,
+  };
+
+  // Keep legacy hand-authored models inline, but update fields for which the
+  // Requesty API is authoritative. New and already-factored files stay based
+  // on shared model metadata.
+  if (authored !== undefined && authored.base_model === undefined) {
+    return { ...authored, ...values, cost, limit, modalities } as SyncedModel;
+  }
+
+  return factorBaseModel(
+    canonical,
+    values,
+    limit,
+  );
+}
diff --git a/packages/core/test/sync.test.ts b/packages/core/test/sync.test.ts
index d9287ebca8..3e84e4c1da 100644
--- a/packages/core/test/sync.test.ts
+++ b/packages/core/test/sync.test.ts
@@ -33,6 +33,11 @@ import {
 } from "../src/sync/providers/openrouter.js";
 import { buildLLMGatewayModel, type LLMGatewayModel } from "../src/sync/providers/llmgateway.js";
 import { openai, parseOpenAIModels } from "../src/sync/providers/openai.js";
+import {
+  buildRequestyModel,
+  requesty,
+  type RequestyModel,
+} from "../src/sync/providers/requesty.js";
 import { resolveVeniceBaseModel } from "../src/sync/providers/venice.js";
 import { buildVercelModel, vercel } from "../src/sync/providers/vercel.js";
 import { buildWandbModel, type WandbModel } from "../src/sync/providers/wandb.js";
@@ -66,6 +71,155 @@ function anthropicModel(overrides: Partial<AnthropicModel> = {}): AnthropicModel
   };
 }
 
+function requestyModel(overrides: Partial<RequestyModel> = {}): RequestyModel {
+  return {
+    id: "openai/gpt-5.5",
+    api: "chat",
+    input_price: 0.000005,
+    output_price: 0.00003,
+    cached_price: 0.0000005,
+    context_window: 1_050_000,
+    max_output_tokens: 128_000,
+    supports_caching: true,
+    supports_vision: true,
+    supports_reasoning: true,
+    supports_tool_calling: true,
+    supports_image_generation: false,
+    supports_output_json_object: true,
+    supports_output_json_schema: true,
+    ...overrides,
+  };
+}
+
+test("Requesty sync includes Responses routes from the chat catalog", () => {
+  const model = requestyModel({ id: "openai-responses/gpt-5.5" });
+  expect(requesty.parseModels({ data: [model] })).toEqual([model]);
+});
+
+test("Requesty sync keeps inherited image output modalities", () => {
+  const model = buildRequestyModel(
+    requestyModel({
+      id: "google/gemini-3.1-flash-image-preview",
+      input_price: 0.0000005,
+      output_price: 0.000002,
+      context_window: 131_072,
+      max_output_tokens: 32_768,
+      supports_image_generation: false,
+    }),
+    {
+      base_model: "google/gemini-3.1-flash-image-preview",
+      modalities: { input: ["text"], output: ["text"] },
+    },
+    { base_model: "google/gemini-3.1-flash-image-preview" },
+    "google/gemini-3.1-flash-image-preview",
+  );
+
+  // The shared metadata already declares image output. Requesty's coarse
+  // capability flag must not replace it with a text-only modality override.
+  expect(model).not.toHaveProperty("modalities");
+});
+
+test("Requesty sync preserves intrinsic reasoning and declares controls", () => {
+  const deepseek = buildRequestyModel(
+    requestyModel({
+      id: "deepseek/deepseek-reasoner",
+      input_price: 0.00000014,
+      output_price: 0.00000028,
+      context_window: 1_000_000,
+      max_output_tokens: 384_000,
+      supports_vision: false,
+      supports_reasoning: false,
+      supports_output_json_schema: false,
+    }),
+    {
+      base_model: "deepseek/deepseek-reasoner",
+      reasoning: false,
+      reasoning_options: [],
+    },
+    { base_model: "deepseek/deepseek-reasoner", reasoning: false },
+    "deepseek/deepseek-reasoner",
+  );
+  expect(deepseek).not.toHaveProperty("reasoning");
+  expect(deepseek).toMatchObject({ reasoning_options: [] });
+
+  const openai = buildRequestyModel(
+    requestyModel({ id: "openai/gpt-5.4-mini" }),
+    undefined,
+    undefined,
+    "openai/gpt-5.4-mini",
+  );
+  expect(openai).toMatchObject({
+    reasoning_options: [
+      { type: "effort", values: ["none", "low", "medium", "high", "max"] },
+      { type: "budget_tokens" },
+    ],
+  });
+
+  for (const [id, canonical] of [
+    ["openai-responses/gpt-5.4-mini", "openai/gpt-5.4-mini"],
+    ["openai-responses/gpt-5.4-nano", "openai/gpt-5.4-nano"],
+    ["openai-responses/gpt-5.5", "openai/gpt-5.5"],
+    ["openai-responses/gpt-5.6-luna", "openai/gpt-5.6-luna"],
+    ["openai-responses/gpt-5.6-sol", "openai/gpt-5.6-sol"],
+    ["openai-responses/gpt-5.6-terra", "openai/gpt-5.6-terra"],
+  ]) {
+    const responsesModel = buildRequestyModel(
+      requestyModel({ id }),
+      undefined,
+      undefined,
+      canonical,
+    );
+    expect(responsesModel).toMatchObject({
+      provider: { npm: "@ai-sdk/openai", shape: "responses" },
+    });
+  }
+
+  const image = buildRequestyModel(
+    requestyModel({ id: "google/gemini-3.1-flash-image-preview" }),
+    undefined,
+    undefined,
+    "google/gemini-3.1-flash-image-preview",
+  );
+  expect(image).not.toHaveProperty("tool_call");
+
+  const gemma = buildRequestyModel(
+    requestyModel({ id: "google/gemma-4-31b-it" }),
+    {
+      base_model: "google/gemma-4-31b-it",
+      reasoning_options: [
+        { type: "effort", values: ["none", "low", "medium", "high", "max"] },
+        { type: "budget_tokens" },
+      ],
+    },
+    { base_model: "google/gemma-4-31b-it" },
+    "google/gemma-4-31b-it",
+  );
+  expect(gemma).toMatchObject({ reasoning_options: [] });
+});
+
+test("Requesty sync updates existing provider-specific values", () => {
+  const model = buildRequestyModel(
+    requestyModel({ context_window: 900_000, max_output_tokens: 64_000 }),
+    {
+      base_model: "openai/gpt-5.5",
+      reasoning: true,
+      reasoning_options: [],
+      cost: { input: 1, output: 2 },
+      limit: { context: 400_000, output: 32_000 },
+      modalities: { input: ["text", "image"], output: ["text"] },
+    },
+    { base_model: "openai/gpt-5.5" },
+    "openai/gpt-5.5",
+  );
+
+  expect(model).toMatchObject({
+    cost: { input: 5, output: 30, cache_read: 0.5 },
+    limit: { context: 900_000, output: 64_000 },
+  });
+  expect("sameModel" in requesty).toBe(false);
+  expect(requesty.deleteMissing).toBe(false);
+});
+
 const anthropicPricingMarkdown = `
 ## Model pricing
 
diff --git a/providers/requesty/models/alibaba/qwen-max.toml b/providers/requesty/models/alibaba/qwen-max.toml
new file mode 100644
index 0000000000..d44732ff85
--- /dev/null
+++ b/providers/requesty/models/alibaba/qwen-max.toml
@@ -0,0 +1,6 @@
+base_model = "alibaba/qwen-max"
+structured_output = true
+
+[cost]
+input = 1.6
+output = 6.4
diff --git a/providers/requesty/models/alibaba/qwen-plus.toml b/providers/requesty/models/alibaba/qwen-plus.toml
new file mode 100644
index 0000000000..6471f13f9d
--- /dev/null
+++ b/providers/requesty/models/alibaba/qwen-plus.toml
@@ -0,0 +1,10 @@
+base_model = "alibaba/qwen-plus"
+structured_output = true
+reasoning_options = []
+
+[cost]
+input = 0.4
+output = 1.2
+
+[limit]
+context = 131_072
diff --git a/providers/requesty/models/alibaba/qwen-turbo.toml b/providers/requesty/models/alibaba/qwen-turbo.toml
new file mode 100644
index 0000000000..1063e96bd5
--- /dev/null
+++ b/providers/requesty/models/alibaba/qwen-turbo.toml
@@ -0,0 +1,7 @@
+base_model = "alibaba/qwen-turbo"
+structured_output = true
+reasoning_options = []
+
+[cost]
+input = 0.05
+output = 0.2
diff --git a/providers/requesty/models/alibaba/qwen3-coder-flash.toml b/providers/requesty/models/alibaba/qwen3-coder-flash.toml
new file mode 100644
index 0000000000..114463be6b
--- /dev/null
+++ b/providers/requesty/models/alibaba/qwen3-coder-flash.toml
@@ -0,0 +1,15 @@
+base_model = "alibaba/qwen3-coder-flash"
+attachment = true
+structured_output = true
+
+[cost]
+input = 0.3
+output = 1.5
+cache_read = 0.08
+cache_write = 0.3
+
+[limit]
+context = 1_048_576
+
+[modalities]
+input = ["text", "image"]
diff --git a/providers/requesty/models/alibaba/qwen3-coder-plus.toml b/providers/requesty/models/alibaba/qwen3-coder-plus.toml
new file mode 100644
index 0000000000..0d54c623d9
--- /dev/null
+++ b/providers/requesty/models/alibaba/qwen3-coder-plus.toml
@@ -0,0 +1,10 @@
+base_model = "alibaba/qwen3-coder-plus"
+attachment = true
+structured_output = true
+
+[cost]
+input = 1
+output = 5
+
+[modalities]
+input = ["text", "image"]
diff --git a/providers/requesty/models/alibaba/qwen3-max.toml b/providers/requesty/models/alibaba/qwen3-max.toml
new file mode 100644
index 0000000000..90b627bf10
--- /dev/null
+++ b/providers/requesty/models/alibaba/qwen3-max.toml
@@ -0,0 +1,10 @@
+base_model = "alibaba/qwen3-max"
+attachment = true
+structured_output = true
+
+[cost]
+input = 0.861
+output = 3.441
+
+[modalities]
+input = ["text", "image"]
diff --git a/providers/requesty/models/alibaba/qwen3.6-plus.toml b/providers/requesty/models/alibaba/qwen3.6-plus.toml
new file mode 100644
index 0000000000..31f43fd9fa
--- /dev/null
+++ b/providers/requesty/models/alibaba/qwen3.6-plus.toml
@@ -0,0 +1,10 @@
+base_model = "alibaba/qwen3.6-plus"
+attachment = true
+structured_output = true
+reasoning_options = []
+
+[cost]
+input = 0.5
+output = 3
+cache_read = 0.05
+cache_write = 0.625
diff --git a/providers/requesty/models/alibaba/qwen3.7-max.toml b/providers/requesty/models/alibaba/qwen3.7-max.toml
new file mode 100644
index 0000000000..22d302d182
--- /dev/null
+++ b/providers/requesty/models/alibaba/qwen3.7-max.toml
@@ -0,0 +1,12 @@
+base_model = "alibaba/qwen3.7-max"
+structured_output = true
+reasoning_options = []
+
+[cost]
+input = 2.5
+output = 7.5
+cache_read = 0.25
+cache_write = 3.125
+
+[limit]
+context = 1_048_576
diff --git a/providers/requesty/models/alibaba/qwen3.7-plus.toml b/providers/requesty/models/alibaba/qwen3.7-plus.toml
new file mode 100644
index 0000000000..e5659e54d1
--- /dev/null
+++ b/providers/requesty/models/alibaba/qwen3.7-plus.toml
@@ -0,0 +1,14 @@
+base_model = "alibaba/qwen3.7-plus"
+attachment = true
+structured_output = true
+reasoning_options = []
+
+[cost]
+input = 0.32
+output = 1.28
+cache_read = 0.032
+cache_write = 0.4
+
+[limit]
+context = 1_048_576
+output = 65_536
diff --git a/providers/requesty/models/anthropic/claude-fable-5.toml b/providers/requesty/models/anthropic/claude-fable-5.toml
new file mode 100644
index 0000000000..685f20fb08
--- /dev/null
+++ b/providers/requesty/models/anthropic/claude-fable-5.toml
@@ -0,0 +1,15 @@
+base_model = "anthropic/claude-fable-5"
+structured_output = true
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 10
+output = 50
+cache_read = 1
+cache_write = 12.5
diff --git a/providers/requesty/models/anthropic/claude-haiku-4-5.toml b/providers/requesty/models/anthropic/claude-haiku-4-5.toml
index d2238e3fb5..01944a248b 100644
--- a/providers/requesty/models/anthropic/claude-haiku-4-5.toml
+++ b/providers/requesty/models/anthropic/claude-haiku-4-5.toml
@@ -5,21 +5,28 @@ release_date = "2025-10-15"
 last_updated = "2025-10-15"
 attachment = true
 reasoning = true
-reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "max"] }, { type = "budget_tokens" }]
 temperature = true
 tool_call = true
+structured_output = true
 knowledge = "2025-02-01"
 open_weights = false
 
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
 [cost]
-input = 1.00
-output = 5.00
+input = 1
+output = 5
 cache_read = 0.1
 cache_write = 1.25
 
 [limit]
 context = 200_000
-output = 62_000
+output = 64_000
 
 [modalities]
 input = ["text", "image", "pdf"]
diff --git a/providers/requesty/models/anthropic/claude-opus-4-1.toml b/providers/requesty/models/anthropic/claude-opus-4-1.toml
index e98d82bf65..410e9e48dd 100644
--- a/providers/requesty/models/anthropic/claude-opus-4-1.toml
+++ b/providers/requesty/models/anthropic/claude-opus-4-1.toml
@@ -1,6 +1,12 @@
 base_model = "anthropic/claude-opus-4-1-20250805"
+structured_output = true
 
-reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "max"] }, { type = "budget_tokens" }]
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
 
 [cost]
 input = 15
diff --git a/providers/requesty/models/anthropic/claude-opus-4-5.toml b/providers/requesty/models/anthropic/claude-opus-4-5.toml
index 8710c9a3c7..7be00beaaf 100644
--- a/providers/requesty/models/anthropic/claude-opus-4-5.toml
+++ b/providers/requesty/models/anthropic/claude-opus-4-5.toml
@@ -5,15 +5,22 @@ release_date = "2025-11-24"
 last_updated = "2025-11-24"
 attachment = true
 reasoning = true
-reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "max"] }, { type = "budget_tokens" }]
 temperature = true
 tool_call = true
+structured_output = true
 knowledge = "2025-03-31"
 open_weights = false
 
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
 [cost]
-input = 5.00
-output = 25.00
+input = 5
+output = 25
 cache_read = 0.5
 cache_write = 6.25
 
diff --git a/providers/requesty/models/anthropic/claude-opus-4-7.toml b/providers/requesty/models/anthropic/claude-opus-4-7.toml
new file mode 100644
index 0000000000..4c464ebf79
--- /dev/null
+++ b/providers/requesty/models/anthropic/claude-opus-4-7.toml
@@ -0,0 +1,15 @@
+base_model = "anthropic/claude-opus-4-7"
+structured_output = true
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 5
+output = 25
+cache_read = 0.5
+cache_write = 6.25
diff --git a/providers/requesty/models/anthropic/claude-opus-4-8.toml b/providers/requesty/models/anthropic/claude-opus-4-8.toml
new file mode 100644
index 0000000000..32911565b0
--- /dev/null
+++ b/providers/requesty/models/anthropic/claude-opus-4-8.toml
@@ -0,0 +1,15 @@
+base_model = "anthropic/claude-opus-4-8"
+structured_output = true
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 5
+output = 25
+cache_read = 0.5
+cache_write = 6.25
diff --git a/providers/requesty/models/anthropic/claude-sonnet-4-5.toml b/providers/requesty/models/anthropic/claude-sonnet-4-5.toml
index 612054daad..0d04a34bea 100644
--- a/providers/requesty/models/anthropic/claude-sonnet-4-5.toml
+++ b/providers/requesty/models/anthropic/claude-sonnet-4-5.toml
@@ -1,6 +1,12 @@
 base_model = "anthropic/claude-sonnet-4-5-20250929"
+structured_output = true
 
-reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "max"] }, { type = "budget_tokens" }]
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
 
 [cost]
 input = 3
diff --git a/providers/requesty/models/anthropic/claude-sonnet-4-6.toml b/providers/requesty/models/anthropic/claude-sonnet-4-6.toml
index 186b6d51c4..508cf2ab74 100644
--- a/providers/requesty/models/anthropic/claude-sonnet-4-6.toml
+++ b/providers/requesty/models/anthropic/claude-sonnet-4-6.toml
@@ -5,30 +5,36 @@ release_date = "2026-02-17"
 last_updated = "2026-02-17"
 attachment = true
 reasoning = true
-reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "max"] }, { type = "budget_tokens" }]
 temperature = true
 tool_call = true
 structured_output = true
 knowledge = "2025-08-31"
 open_weights = false
 
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
 [cost]
-input = 3.00
-output = 15.00
-cache_read = 0.30
+input = 3
+output = 15
+cache_read = 0.3
 cache_write = 3.75
 
 [[cost.tiers]]
-tier = { size = 200_000 }
-input = 6.00
-output = 22.50
-cache_read = 0.60
-cache_write = 7.50
+tier = { type = "context", size = 200_000 }
+input = 6
+output = 22.5
+cache_read = 0.6
+cache_write = 7.5
 
 [limit]
 context = 1_000_000
-output = 128_000
+output = 64_000
 
 [modalities]
-input = ["text", "image"]
+input = ["text", "image", "pdf"]
 output = ["text"]
diff --git a/providers/requesty/models/anthropic/claude-sonnet-5.toml b/providers/requesty/models/anthropic/claude-sonnet-5.toml
new file mode 100644
index 0000000000..ef08197229
--- /dev/null
+++ b/providers/requesty/models/anthropic/claude-sonnet-5.toml
@@ -0,0 +1,15 @@
+base_model = "anthropic/claude-sonnet-5"
+structured_output = true
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 2
+output = 10
+cache_read = 0.2
+cache_write = 2.5
diff --git a/providers/requesty/models/deepseek/deepseek-chat.toml b/providers/requesty/models/deepseek/deepseek-chat.toml
new file mode 100644
index 0000000000..e07c492dbf
--- /dev/null
+++ b/providers/requesty/models/deepseek/deepseek-chat.toml
@@ -0,0 +1,8 @@
+base_model = "deepseek/deepseek-chat"
+attachment = false
+structured_output = true
+
+[cost]
+input = 0.14
+output = 0.28
+cache_read = 0.028
diff --git a/providers/requesty/models/deepseek/deepseek-reasoner.toml b/providers/requesty/models/deepseek/deepseek-reasoner.toml
new file mode 100644
index 0000000000..f286d0cde0
--- /dev/null
+++ b/providers/requesty/models/deepseek/deepseek-reasoner.toml
@@ -0,0 +1,9 @@
+base_model = "deepseek/deepseek-reasoner"
+attachment = false
+structured_output = true
+reasoning_options = []
+
+[cost]
+input = 0.14
+output = 0.28
+cache_read = 0.028
diff --git a/providers/requesty/models/deepseek/deepseek-v4-flash.toml b/providers/requesty/models/deepseek/deepseek-v4-flash.toml
new file mode 100644
index 0000000000..16e3dfec13
--- /dev/null
+++ b/providers/requesty/models/deepseek/deepseek-v4-flash.toml
@@ -0,0 +1,7 @@
+base_model = "deepseek/deepseek-v4-flash"
+reasoning_options = []
+
+[cost]
+input = 0.14
+output = 0.28
+cache_read = 0.0028
diff --git a/providers/requesty/models/deepseek/deepseek-v4-pro.toml b/providers/requesty/models/deepseek/deepseek-v4-pro.toml
new file mode 100644
index 0000000000..bd62eeb744
--- /dev/null
+++ b/providers/requesty/models/deepseek/deepseek-v4-pro.toml
@@ -0,0 +1,7 @@
+base_model = "deepseek/deepseek-v4-pro"
+reasoning_options = []
+
+[cost]
+input = 0.435
+output = 0.87
+cache_read = 0.0145
diff --git a/providers/requesty/models/google/gemini-2.5-flash-lite.toml b/providers/requesty/models/google/gemini-2.5-flash-lite.toml
new file mode 100644
index 0000000000..47ac3bff5a
--- /dev/null
+++ b/providers/requesty/models/google/gemini-2.5-flash-lite.toml
@@ -0,0 +1,16 @@
+base_model = "google/gemini-2.5-flash-lite"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 0.1
+output = 0.4
+cache_read = 0.01
+
+[limit]
+output = 65_535
diff --git a/providers/requesty/models/google/gemini-2.5-flash.toml b/providers/requesty/models/google/gemini-2.5-flash.toml
index c341820f3e..f92e0a0a62 100644
--- a/providers/requesty/models/google/gemini-2.5-flash.toml
+++ b/providers/requesty/models/google/gemini-2.5-flash.toml
@@ -5,21 +5,28 @@ release_date = "2025-06-17"
 last_updated = "2025-06-17"
 attachment = true
 reasoning = true
-reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "max"] }, { type = "budget_tokens" }]
 temperature = true
-knowledge = "2025-01"
 tool_call = true
+structured_output = true
+knowledge = "2025-01"
 open_weights = false
 
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
 [cost]
-input = 0.30
-output = 2.50
+input = 0.3
+output = 2.5
 cache_read = 0.075
 cache_write = 0.55
 
 [limit]
 context = 1_048_576
-output = 65_536
+output = 65_535
 
 [modalities]
 input = ["text", "image", "audio", "video", "pdf"]
diff --git a/providers/requesty/models/google/gemini-2.5-pro.toml b/providers/requesty/models/google/gemini-2.5-pro.toml
index 87bcb7dff5..9c09fa539f 100644
--- a/providers/requesty/models/google/gemini-2.5-pro.toml
+++ b/providers/requesty/models/google/gemini-2.5-pro.toml
@@ -1,7 +1,12 @@
 base_model = "google/gemini-2.5-pro"
 base_model_omit = ["structured_output"]
 
-reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "max"] }, { type = "budget_tokens" }]
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
 
 [cost]
 input = 1.25
@@ -14,3 +19,6 @@ tier = { type = "context", size = 200_000 }
 input = 2.5
 output = 15
 cache_read = 0.25
+
+[limit]
+output = 65_535
diff --git a/providers/requesty/models/google/gemini-3-flash-preview.toml b/providers/requesty/models/google/gemini-3-flash-preview.toml
index ca138cdb2a..cdf3bf3343 100644
--- a/providers/requesty/models/google/gemini-3-flash-preview.toml
+++ b/providers/requesty/models/google/gemini-3-flash-preview.toml
@@ -4,12 +4,19 @@ release_date = "2025-12-17"
 last_updated = "2025-12-17"
 attachment = true
 reasoning = true
-reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "max"] }, { type = "budget_tokens" }]
 temperature = true
-knowledge = "2025-01"
 tool_call = true
+structured_output = true
+knowledge = "2025-01"
 open_weights = false
 
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
 [cost]
 input = 0.5
 output = 3
@@ -18,7 +25,7 @@ cache_write = 1
 
 [limit]
 context = 1_048_576
-output = 65_536
+output = 65_535
 
 [modalities]
 input = ["text", "image", "audio", "video", "pdf"]
diff --git a/providers/requesty/models/google/gemini-3-pro-preview.toml b/providers/requesty/models/google/gemini-3-pro-preview.toml
index 037e789bfc..21dacadb44 100644
--- a/providers/requesty/models/google/gemini-3-pro-preview.toml
+++ b/providers/requesty/models/google/gemini-3-pro-preview.toml
@@ -5,21 +5,28 @@ release_date = "2025-11-18"
 last_updated = "2025-11-18"
 attachment = true
 reasoning = true
-reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "max"] }, { type = "budget_tokens" }]
 temperature = true
-knowledge = "2025-01"
 tool_call = true
+structured_output = true
+knowledge = "2025-01"
 open_weights = false
 
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
 [cost]
-input = 2.00
-output = 12.00
+input = 2
+output = 12
 cache_read = 0.2
-cache_write = 4.50
+cache_write = 4.5
 
 [limit]
 context = 1_048_576
-output = 65_536
+output = 65_535
 
 [modalities]
 input = ["text", "image", "audio", "video", "pdf"]
diff --git a/providers/requesty/models/google/gemini-3.1-flash-image-preview.toml b/providers/requesty/models/google/gemini-3.1-flash-image-preview.toml
new file mode 100644
index 0000000000..d3f3ff82b3
--- /dev/null
+++ b/providers/requesty/models/google/gemini-3.1-flash-image-preview.toml
@@ -0,0 +1,19 @@
+base_model = "google/gemini-3.1-flash-image-preview"
+structured_output = false
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 0.5
+output = 2
+cache_read = 0
+cache_write = 0
+
+[limit]
+context = 131_072
+output = 32_768
diff --git a/providers/requesty/models/google/gemini-3.1-flash-lite-preview.toml b/providers/requesty/models/google/gemini-3.1-flash-lite-preview.toml
new file mode 100644
index 0000000000..c1e9c12bbb
--- /dev/null
+++ b/providers/requesty/models/google/gemini-3.1-flash-lite-preview.toml
@@ -0,0 +1,16 @@
+base_model = "google/gemini-3.1-flash-lite-preview"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 0.25
+output = 1.5
+cache_read = 0.025
+
+[limit]
+output = 65_535
diff --git a/providers/requesty/models/google/gemini-3.1-pro-preview.toml b/providers/requesty/models/google/gemini-3.1-pro-preview.toml
new file mode 100644
index 0000000000..15bcd070cf
--- /dev/null
+++ b/providers/requesty/models/google/gemini-3.1-pro-preview.toml
@@ -0,0 +1,17 @@
+base_model = "google/gemini-3.1-pro-preview"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 2
+output = 12
+cache_read = 0.2
+cache_write = 4.5
+
+[limit]
+output = 65_535
diff --git a/providers/requesty/models/google/gemma-4-31b-it.toml b/providers/requesty/models/google/gemma-4-31b-it.toml
new file mode 100644
index 0000000000..7d7fc64cce
--- /dev/null
+++ b/providers/requesty/models/google/gemma-4-31b-it.toml
@@ -0,0 +1,10 @@
+base_model = "google/gemma-4-31b-it"
+structured_output = false
+reasoning_options = []
+
+[cost]
+input = 0
+output = 0
+
+[limit]
+output = 8_192
diff --git a/providers/requesty/models/nvidia/nemotron-3-nano-30b-a3b.toml b/providers/requesty/models/nvidia/nemotron-3-nano-30b-a3b.toml
new file mode 100644
index 0000000000..69ce06e8b1
--- /dev/null
+++ b/providers/requesty/models/nvidia/nemotron-3-nano-30b-a3b.toml
@@ -0,0 +1,7 @@
+base_model = "nvidia/nemotron-3-nano-30b-a3b"
+structured_output = false
+reasoning_options = []
+
+[cost]
+input = 0
+output = 0
diff --git a/providers/requesty/models/nvidia/nemotron-3-nano-omni-30b-a3b-reasoning.toml b/providers/requesty/models/nvidia/nemotron-3-nano-omni-30b-a3b-reasoning.toml
new file mode 100644
index 0000000000..7bb9a33108
--- /dev/null
+++ b/providers/requesty/models/nvidia/nemotron-3-nano-omni-30b-a3b-reasoning.toml
@@ -0,0 +1,11 @@
+base_model = "nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"
+structured_output = false
+reasoning_options = []
+
+[cost]
+input = 0
+output = 0
+
+[limit]
+context = 131_072
+output = 20_480
diff --git a/providers/requesty/models/nvidia/nemotron-3-super-120b-a12b.toml b/providers/requesty/models/nvidia/nemotron-3-super-120b-a12b.toml
new file mode 100644
index 0000000000..3069ade1dd
--- /dev/null
+++ b/providers/requesty/models/nvidia/nemotron-3-super-120b-a12b.toml
@@ -0,0 +1,11 @@
+base_model = "nvidia/nemotron-3-super-120b-a12b"
+structured_output = false
+reasoning_options = []
+
+[cost]
+input = 0
+output = 0
+
+[limit]
+context = 1_048_576
+output = 65_536
diff --git a/providers/requesty/models/nvidia/nemotron-3-ultra-550b-a55b.toml b/providers/requesty/models/nvidia/nemotron-3-ultra-550b-a55b.toml
new file mode 100644
index 0000000000..9bbbcda0cb
--- /dev/null
+++ b/providers/requesty/models/nvidia/nemotron-3-ultra-550b-a55b.toml
@@ -0,0 +1,11 @@
+base_model = "nvidia/nemotron-3-ultra-550b-a55b"
+structured_output = false
+reasoning_options = []
+
+[cost]
+input = 0
+output = 0
+
+[limit]
+context = 1_048_576
+output = 65_536
diff --git a/providers/requesty/models/openai-responses/gpt-5.4-mini.toml b/providers/requesty/models/openai-responses/gpt-5.4-mini.toml
new file mode 100644
index 0000000000..81e1240e2d
--- /dev/null
+++ b/providers/requesty/models/openai-responses/gpt-5.4-mini.toml
@@ -0,0 +1,17 @@
+base_model = "openai/gpt-5.4-mini"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 0.75
+output = 4.5
+cache_read = 0.075
+
+[provider]
+npm = "@ai-sdk/openai"
+shape = "responses"
diff --git a/providers/requesty/models/openai-responses/gpt-5.4-nano.toml b/providers/requesty/models/openai-responses/gpt-5.4-nano.toml
new file mode 100644
index 0000000000..380d77ab46
--- /dev/null
+++ b/providers/requesty/models/openai-responses/gpt-5.4-nano.toml
@@ -0,0 +1,17 @@
+base_model = "openai/gpt-5.4-nano"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 0.2
+output = 1.25
+cache_read = 0.02
+
+[provider]
+npm = "@ai-sdk/openai"
+shape = "responses"
diff --git a/providers/requesty/models/openai-responses/gpt-5.5.toml b/providers/requesty/models/openai-responses/gpt-5.5.toml
new file mode 100644
index 0000000000..dff17b35af
--- /dev/null
+++ b/providers/requesty/models/openai-responses/gpt-5.5.toml
@@ -0,0 +1,17 @@
+base_model = "openai/gpt-5.5"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 5
+output = 30
+cache_read = 0.5
+
+[provider]
+npm = "@ai-sdk/openai"
+shape = "responses"
diff --git a/providers/requesty/models/openai-responses/gpt-5.6-luna.toml b/providers/requesty/models/openai-responses/gpt-5.6-luna.toml
new file mode 100644
index 0000000000..153b58b13a
--- /dev/null
+++ b/providers/requesty/models/openai-responses/gpt-5.6-luna.toml
@@ -0,0 +1,17 @@
+base_model = "openai/gpt-5.6-luna"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 1
+output = 6
+cache_read = 0.1
+
+[provider]
+npm = "@ai-sdk/openai"
+shape = "responses"
diff --git a/providers/requesty/models/openai-responses/gpt-5.6-sol.toml b/providers/requesty/models/openai-responses/gpt-5.6-sol.toml
new file mode 100644
index 0000000000..ef8bf3f0d3
--- /dev/null
+++ b/providers/requesty/models/openai-responses/gpt-5.6-sol.toml
@@ -0,0 +1,17 @@
+base_model = "openai/gpt-5.6-sol"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 5
+output = 30
+cache_read = 0.5
+
+[provider]
+npm = "@ai-sdk/openai"
+shape = "responses"
diff --git a/providers/requesty/models/openai-responses/gpt-5.6-terra.toml b/providers/requesty/models/openai-responses/gpt-5.6-terra.toml
new file mode 100644
index 0000000000..312c651a40
--- /dev/null
+++ b/providers/requesty/models/openai-responses/gpt-5.6-terra.toml
@@ -0,0 +1,17 @@
+base_model = "openai/gpt-5.6-terra"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 2.5
+output = 15
+cache_read = 0.25
+
+[provider]
+npm = "@ai-sdk/openai"
+shape = "responses"
diff --git a/providers/requesty/models/openai/gpt-4.1-mini.toml b/providers/requesty/models/openai/gpt-4.1-mini.toml
index 5b1b2499bc..cf48501957 100644
--- a/providers/requesty/models/openai/gpt-4.1-mini.toml
+++ b/providers/requesty/models/openai/gpt-4.1-mini.toml
@@ -7,18 +7,19 @@ attachment = true
 reasoning = false
 temperature = true
 tool_call = true
+structured_output = true
 knowledge = "2024-04"
 open_weights = false
 
 [cost]
-input = 0.40
-output = 1.60
-cache_read = 0.10
+input = 0.4
+output = 1.6
+cache_read = 0.1
 
 [limit]
 context = 1_047_576
 output = 32_768
 
 [modalities]
-input = ["text", "image"]
+input = ["text", "image", "pdf"]
 output = ["text"]
diff --git a/providers/requesty/models/openai/gpt-4.1-nano.toml b/providers/requesty/models/openai/gpt-4.1-nano.toml
new file mode 100644
index 0000000000..f85c094770
--- /dev/null
+++ b/providers/requesty/models/openai/gpt-4.1-nano.toml
@@ -0,0 +1,6 @@
+base_model = "openai/gpt-4.1-nano"
+
+[cost]
+input = 0.1
+output = 0.4
+cache_read = 0.025
diff --git a/providers/requesty/models/openai/gpt-4.1.toml b/providers/requesty/models/openai/gpt-4.1.toml
index 67e617230a..0501119e22 100644
--- a/providers/requesty/models/openai/gpt-4.1.toml
+++ b/providers/requesty/models/openai/gpt-4.1.toml
@@ -5,6 +5,3 @@ base_model_omit = ["structured_output"]
 input = 2
 output = 8
 cache_read = 0.5
-
-[modalities]
-input = ["text", "image"]
diff --git a/providers/requesty/models/openai/gpt-4o-2024-05-13.toml b/providers/requesty/models/openai/gpt-4o-2024-05-13.toml
new file mode 100644
index 0000000000..d2d5d312e7
--- /dev/null
+++ b/providers/requesty/models/openai/gpt-4o-2024-05-13.toml
@@ -0,0 +1,7 @@
+base_model = "openai/gpt-4o-2024-05-13"
+structured_output = false
+
+[cost]
+input = 2.5
+output = 10
+cache_read = 2.5
diff --git a/providers/requesty/models/openai/gpt-4o-2024-08-06.toml b/providers/requesty/models/openai/gpt-4o-2024-08-06.toml
new file mode 100644
index 0000000000..344cd161a7
--- /dev/null
+++ b/providers/requesty/models/openai/gpt-4o-2024-08-06.toml
@@ -0,0 +1,6 @@
+base_model = "openai/gpt-4o-2024-08-06"
+
+[cost]
+input = 2.5
+output = 10
+cache_read = 1.25
diff --git a/providers/requesty/models/openai/gpt-4o-2024-11-20.toml b/providers/requesty/models/openai/gpt-4o-2024-11-20.toml
new file mode 100644
index 0000000000..c807afdbe1
--- /dev/null
+++ b/providers/requesty/models/openai/gpt-4o-2024-11-20.toml
@@ -0,0 +1,6 @@
+base_model = "openai/gpt-4o-2024-11-20"
+
+[cost]
+input = 2.5
+output = 10
+cache_read = 1.25
diff --git a/providers/requesty/models/openai/gpt-4o-mini.toml b/providers/requesty/models/openai/gpt-4o-mini.toml
index cbf5bc6092..86a921f094 100644
--- a/providers/requesty/models/openai/gpt-4o-mini.toml
+++ b/providers/requesty/models/openai/gpt-4o-mini.toml
@@ -7,18 +7,19 @@ attachment = true
 reasoning = false
 temperature = true
 tool_call = true
+structured_output = true
 knowledge = "2024-10"
 open_weights = false
 
 [cost]
 input = 0.15
-output = 0.60
-cache_read = 0.08
+output = 0.6
+cache_read = 0.075
 
 [limit]
 context = 128_000
 output = 16_384
 
 [modalities]
-input = ["text", "image"]
+input = ["text", "image", "pdf"]
 output = ["text"]
diff --git a/providers/requesty/models/openai/gpt-4o.toml b/providers/requesty/models/openai/gpt-4o.toml
new file mode 100644
index 0000000000..1b57e52012
--- /dev/null
+++ b/providers/requesty/models/openai/gpt-4o.toml
@@ -0,0 +1,6 @@
+base_model = "openai/gpt-4o"
+
+[cost]
+input = 2.5
+output = 10
+cache_read = 1.25
diff --git a/providers/requesty/models/openai/gpt-5-mini.toml b/providers/requesty/models/openai/gpt-5-mini.toml
index 62e297e5c2..eaf3473843 100644
--- a/providers/requesty/models/openai/gpt-5-mini.toml
+++ b/providers/requesty/models/openai/gpt-5-mini.toml
@@ -5,20 +5,28 @@ release_date = "2025-08-07"
 last_updated = "2025-08-07"
 attachment = true
 reasoning = true
-reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "max"] }, { type = "budget_tokens" }]
 temperature = false
-knowledge = "2024-05-30"
 tool_call = true
+structured_output = true
+knowledge = "2024-05-30"
 open_weights = false
 
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
 [cost]
 input = 0.25
-output = 2.00
-cache_read = 0.03
+output = 2
+cache_read = 0.025
 
 [limit]
-context = 128_000
-output = 32_000
+context = 400_000
+input = 272_000
+output = 128_000
 
 [modalities]
 input = ["text", "image"]
diff --git a/providers/requesty/models/openai/gpt-5-nano.toml b/providers/requesty/models/openai/gpt-5-nano.toml
index 316044c999..92ea325e90 100644
--- a/providers/requesty/models/openai/gpt-5-nano.toml
+++ b/providers/requesty/models/openai/gpt-5-nano.toml
@@ -5,21 +5,29 @@ release_date = "2025-08-07"
 last_updated = "2025-08-07"
 attachment = true
 reasoning = true
-reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "max"] }, { type = "budget_tokens" }]
 temperature = false
-knowledge = "2024-05-30"
 tool_call = true
+structured_output = true
+knowledge = "2024-05-30"
 open_weights = false
 
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
 [cost]
 input = 0.05
-output = 0.40
-cache_read = 0.01
+output = 0.4
+cache_read = 0.005
 
 [limit]
-context = 16_000
-output = 4_000
+context = 400_000
+input = 272_000
+output = 128_000
 
 [modalities]
-input = ["text"]
+input = ["text", "image"]
 output = ["text"]
diff --git a/providers/requesty/models/openai/gpt-5.1.toml b/providers/requesty/models/openai/gpt-5.1.toml
index d2c6bc22e6..344ea9f9dc 100644
--- a/providers/requesty/models/openai/gpt-5.1.toml
+++ b/providers/requesty/models/openai/gpt-5.1.toml
@@ -5,20 +5,27 @@ release_date = "2025-11-13"
 last_updated = "2025-11-13"
 attachment = true
 reasoning = true
-reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "max"] }, { type = "budget_tokens" }]
 temperature = true
-knowledge = "2024-09-30"
 tool_call = true
 structured_output = true
+knowledge = "2024-09-30"
 open_weights = false
 
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
 [cost]
 input = 1.25
-output = 10.00
+output = 10
 cache_read = 0.125
 
 [limit]
 context = 400_000
+input = 272_000
 output = 128_000
 
 [modalities]
diff --git a/providers/requesty/models/openai/gpt-5.toml b/providers/requesty/models/openai/gpt-5.toml
index 390940407f..644f0ae61c 100644
--- a/providers/requesty/models/openai/gpt-5.toml
+++ b/providers/requesty/models/openai/gpt-5.toml
@@ -5,19 +5,27 @@ release_date = "2025-08-07"
 last_updated = "2025-08-07"
 attachment = true
 reasoning = true
-reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "max"] }, { type = "budget_tokens" }]
 temperature = false
-knowledge = "2024-09-30"
 tool_call = true
+structured_output = true
+knowledge = "2024-09-30"
 open_weights = false
 
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
 [cost]
 input = 1.25
-output = 10.00
-cache_read = 0.13
+output = 10
+cache_read = 0.125
 
 [limit]
 context = 400_000
+input = 272_000
 output = 128_000
 
 [modalities]
diff --git a/providers/requesty/models/openai/o1.toml b/providers/requesty/models/openai/o1.toml
new file mode 100644
index 0000000000..3d902b37b4
--- /dev/null
+++ b/providers/requesty/models/openai/o1.toml
@@ -0,0 +1,13 @@
+base_model = "openai/o1"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 15
+output = 60
+cache_read = 7.5
diff --git a/providers/requesty/models/openai/o3-mini.toml b/providers/requesty/models/openai/o3-mini.toml
new file mode 100644
index 0000000000..43618ef6ae
--- /dev/null
+++ b/providers/requesty/models/openai/o3-mini.toml
@@ -0,0 +1,13 @@
+base_model = "openai/o3-mini"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 1.1
+output = 4.4
+cache_read = 0.55
diff --git a/providers/requesty/models/openai/o3.toml b/providers/requesty/models/openai/o3.toml
new file mode 100644
index 0000000000..3a1118eec9
--- /dev/null
+++ b/providers/requesty/models/openai/o3.toml
@@ -0,0 +1,13 @@
+base_model = "openai/o3"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 2
+output = 8
+cache_read = 0.5
diff --git a/providers/requesty/models/openai/o4-mini.toml b/providers/requesty/models/openai/o4-mini.toml
index b6f4f9b660..836207d921 100644
--- a/providers/requesty/models/openai/o4-mini.toml
+++ b/providers/requesty/models/openai/o4-mini.toml
@@ -5,16 +5,23 @@ release_date = "2025-04-16"
 last_updated = "2025-04-16"
 attachment = true
 reasoning = true
-reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "max"] }, { type = "budget_tokens" }]
 temperature = true
 tool_call = true
+structured_output = true
 knowledge = "2024-06"
 open_weights = false
 
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+
 [cost]
-input = 1.10
-output = 4.40
-cache_read = 0.28
+input = 1.1
+output = 4.4
+cache_read = 0.275
 
 [limit]
 context = 200_000
diff --git a/providers/requesty/models/xai/grok-4.3.toml b/providers/requesty/models/xai/grok-4.3.toml
new file mode 100644
index 0000000000..2202f0cc7f
--- /dev/null
+++ b/providers/requesty/models/xai/grok-4.3.toml
@@ -0,0 +1,8 @@
+base_model = "xai/grok-4.3"
+reasoning_options = []
+
+[cost]
+input = 1.25
+output = 2.5
+cache_read = 0.2
+cache_write = 1.25
diff --git a/providers/requesty/models/xai/grok-4.5.toml b/providers/requesty/models/xai/grok-4.5.toml
new file mode 100644
index 0000000000..b107b86b7a
--- /dev/null
+++ b/providers/requesty/models/xai/grok-4.5.toml
@@ -0,0 +1,8 @@
+base_model = "xai/grok-4.5"
+reasoning_options = []
+
+[cost]
+input = 2
+output = 6
+cache_read = 0.5
+cache_write = 2
diff --git a/providers/requesty/models/xai/grok-build-0.1.toml b/providers/requesty/models/xai/grok-build-0.1.toml
new file mode 100644
index 0000000000..fc1b5417d5
--- /dev/null
+++ b/providers/requesty/models/xai/grok-build-0.1.toml
@@ -0,0 +1,7 @@
+base_model = "xai/grok-build-0.1"
+reasoning_options = []
+
+[cost]
+input = 1
+output = 2
+cache_read = 0.1
diff --git a/providers/requesty/models/xiaomi/mimo-v2.5-pro.toml b/providers/requesty/models/xiaomi/mimo-v2.5-pro.toml
new file mode 100644
index 0000000000..8756f262ac
--- /dev/null
+++ b/providers/requesty/models/xiaomi/mimo-v2.5-pro.toml
@@ -0,0 +1,8 @@
+base_model = "xiaomi/mimo-v2.5-pro"
+structured_output = false
+reasoning_options = []
+
+[cost]
+input = 0.435
+output = 0.87
+cache_read = 0.0036
diff --git a/providers/requesty/models/xiaomi/mimo-v2.5.toml b/providers/requesty/models/xiaomi/mimo-v2.5.toml
new file mode 100644
index 0000000000..3881b6be3a
--- /dev/null
+++ b/providers/requesty/models/xiaomi/mimo-v2.5.toml
@@ -0,0 +1,8 @@
+base_model = "xiaomi/mimo-v2.5"
+structured_output = false
+reasoning_options = []
+
+[cost]
+input = 0.14
+output = 0.28
+cache_read = 0.0028
diff --git a/providers/requesty/models/zai/glm-5.1.toml b/providers/requesty/models/zai/glm-5.1.toml
new file mode 100644
index 0000000000..852f77a689
--- /dev/null
+++ b/providers/requesty/models/zai/glm-5.1.toml
@@ -0,0 +1,15 @@
+base_model = "zhipuai/glm-5.1"
+attachment = true
+structured_output = false
+reasoning_options = []
+
+[cost]
+input = 1.4
+output = 4.4
+cache_read = 0.26
+
+[limit]
+output = 128_000
+
+[modalities]
+input = ["text", "image"]
diff --git a/providers/requesty/models/zai/glm-5.2.toml b/providers/requesty/models/zai/glm-5.2.toml
new file mode 100644
index 0000000000..4ee091b3be
--- /dev/null
+++ b/providers/requesty/models/zai/glm-5.2.toml
@@ -0,0 +1,15 @@
+base_model = "zhipuai/glm-5.2"
+attachment = true
+structured_output = false
+reasoning_options = []
+
+[cost]
+input = 1.4
+output = 4.4
+cache_read = 0.26
+
+[limit]
+output = 128_000
+
+[modalities]
+input = ["text", "image"]