feat(models): add per-model timeout disable to avoid global override for long-running models (e.g., gpt-5-pro)

- Introduce ModelInfo.disableTimeout to opt out of request timeouts on a per-model basis
- Apply in OpenAI-compatible, Ollama, and LM Studio providers (timeout=0 when flag is true)
- Preserve global “API Request Timeout” behavior (0 still disables globally); per-model flag takes precedence for that model
- Motivation: gpt-5-pro often requires longer runtimes; per-model override avoids forcing a global setting that impacts all models
- Add/extend unit tests to validate provider behavior

packages/types/src/model.ts

@@ -86,6 +86,8 @@ export const modelInfoSchema = z.object({
 	// Capability flag to indicate whether the model supports temperature parameter
 	supportsTemperature: z.boolean().optional(),
 	defaultTemperature: z.number().optional(),
+	// When true, force-disable request timeouts for this model (providers will set timeout=0)
+	disableTimeout: z.boolean().optional(),
 	requiredReasoningBudget: z.boolean().optional(),
 	supportsReasoningEffort: z
 		.union([z.boolean(), z.array(z.enum(["disable", "none", "minimal", "low", "medium", "high", "xhigh"]))])

src/api/providers/__tests__/lm-studio-timeout.spec.ts

@@ -88,4 +88,34 @@ describe("LmStudioHandler timeout configuration", () => {
 			}),
 		)
 	})
+
+	it("should force zero timeout when model info disables timeout", () => {
+		;(getApiRequestTimeout as any).mockReturnValue(600000)
+
+		const spy = vitest.spyOn(LmStudioHandler.prototype as any, "getModel").mockReturnValue({
+			id: "llama2",
+			info: {
+				maxTokens: -1,
+				contextWindow: 128000,
+				supportsPromptCache: false,
+				supportsImages: true,
+				disableTimeout: true,
+			},
+		})
+
+		const options: ApiHandlerOptions = {
+			apiModelId: "llama2",
+			lmStudioModelId: "llama2",
+		}
+
+		new LmStudioHandler(options)
+
+		expect(mockOpenAIConstructor).toHaveBeenCalledWith(
+			expect.objectContaining({
+				timeout: 0,
+			}),
+		)
+
+		spy.mockRestore()
+	})
 })

src/api/providers/__tests__/openai-timeout.spec.ts

@@ -141,4 +141,28 @@ describe("OpenAiHandler timeout configuration", () => {
 			}),
 		)
 	})
+
+	it("should force zero timeout when model info disables timeout", () => {
+		;(getApiRequestTimeout as any).mockReturnValue(600000)
+
+		const options: ApiHandlerOptions = {
+			apiModelId: "gpt-4",
+			openAiModelId: "gpt-4",
+			openAiCustomModelInfo: {
+				maxTokens: -1,
+				contextWindow: 128000,
+				supportsPromptCache: false,
+				supportsImages: true,
+				disableTimeout: true,
+			} as any,
+		}
+
+		new OpenAiHandler(options)
+
+		expect(mockOpenAIConstructor).toHaveBeenCalledWith(
+			expect.objectContaining({
+				timeout: 0, // Forced no timeout via model info
+			}),
+		)
+	})
 })

src/api/providers/lm-studio.ts

@@ -32,7 +32,7 @@ export class LmStudioHandler extends BaseProvider implements SingleCompletionHan
 		this.client = new OpenAI({
 			baseURL: (this.options.lmStudioBaseUrl || "http://localhost:1234") + "/v1",
 			apiKey: apiKey,
-			timeout: getApiRequestTimeout(),
+			timeout: this.getModel().info?.disableTimeout === true ? 0 : getApiRequestTimeout(),
 		})
 	}
 

src/api/providers/openai.ts

@@ -49,7 +49,13 @@ export class OpenAiHandler extends BaseProvider implements SingleCompletionHandl
 			...(this.options.openAiHeaders || {}),
 		}
 
-		const timeout = getApiRequestTimeout()
+		let timeout = getApiRequestTimeout()
+		try {
+			const modelInfo = this.getModel().info
+			if (modelInfo?.disableTimeout === true) {
+				timeout = 0
+			}
+		} catch {}
 
 		if (isAzureAiInference) {
 			// Azure AI Inference Service (e.g., for DeepSeek) uses a different path structure