fix(openai): preserve reasoning for local OpenAI-compatible models via R1 toggle

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

@@ -1118,6 +1118,46 @@ describe("OpenAiHandler", () => {
 			expect(model.id).toBe("")
 			expect(model.info).toBeDefined()
 		})
+
+		it("should set preserveReasoning when openAiR1FormatEnabled is on", () => {
+			const r1Handler = new OpenAiHandler({
+				...mockOptions,
+				openAiR1FormatEnabled: true,
+			})
+			const model = r1Handler.getModel()
+			expect(model.info).toEqual({ ...openAiModelInfoSaneDefaults, preserveReasoning: true })
+		})
+
+		it("should set preserveReasoning for deepseek-reasoner model ids without the toggle", () => {
+			const deepseekHandler = new OpenAiHandler({
+				...mockOptions,
+				openAiModelId: "deepseek-reasoner",
+			})
+			const model = deepseekHandler.getModel()
+			expect(model.info.preserveReasoning).toBe(true)
+		})
+
+		it("should merge preserveReasoning into custom model info when openAiR1FormatEnabled is on", () => {
+			const customInfo: ModelInfo = {
+				...openAiModelInfoSaneDefaults,
+				contextWindow: 32_768,
+				supportsImages: false,
+			}
+			const r1Handler = new OpenAiHandler({
+				...mockOptions,
+				openAiCustomModelInfo: customInfo,
+				openAiR1FormatEnabled: true,
+			})
+			const model = r1Handler.getModel()
+			expect(model.info).toEqual({ ...customInfo, preserveReasoning: true })
+			expect(model.info.contextWindow).toBe(32_768)
+			expect(model.info.supportsImages).toBe(false)
+		})
+
+		it("should not set preserveReasoning by default", () => {
+			const model = handler.getModel()
+			expect(model.info.preserveReasoning).toBeUndefined()
+		})
 	})
 
 	describe("Azure AI Inference Service", () => {

src/api/providers/openai.ts

@@ -98,7 +98,7 @@ export class OpenAiHandler extends BaseProvider implements SingleCompletionHandl
 		// the same strict value as `??`.
 		const strictToolSchemas = this.options.openAiStrictToolSchemas ?? DEFAULT_OPEN_AI_STRICT_TOOL_SCHEMAS
 		const isAzureAiInference = this._isAzureAiInference(modelUrl)
-		const deepseekReasoner = modelId.includes("deepseek-reasoner") || enabledR1Format
+		const deepseekReasoner = this.usesR1Format(modelId)
 
 		if (modelId.includes("o1") || modelId.includes("o3") || modelId.includes("o4")) {
 			yield* this.handleO3FamilyMessage(modelId, systemPrompt, messages, metadata)
@@ -299,6 +299,14 @@ export class OpenAiHandler extends BaseProvider implements SingleCompletionHandl
 		}
 	}
 
+	// Single gate for everything that depends on the R1 request format: message
+	// conversion in createMessage() and reasoning retention in getModel(). Both
+	// must agree, otherwise reasoning is converted for the request but stripped
+	// from the follow-up context (or vice versa).
+	private usesR1Format(modelId: string): boolean {
+		return modelId.includes("deepseek-reasoner") || (this.options.openAiR1FormatEnabled ?? false)
+	}
+
 	override getModel() {
 		const id = this.options.openAiModelId ?? ""
 		const info: ModelInfo = this.options.openAiCustomModelInfo ?? openAiModelInfoSaneDefaults
@@ -311,7 +319,16 @@ export class OpenAiHandler extends BaseProvider implements SingleCompletionHandl
 			settings: { ...this.options, reasoningEffort: info.reasoningEffort },
 			defaultTemperature: 0,
 		})
-		return { id, info, ...params }
+		// Local OpenAI-compatible reasoning models (llama.cpp, LM Studio, Ollama)
+		// stream reasoning_content, but Zoo Code strips it from the follow-up context
+		// unless info.preserveReasoning is set. Whenever createMessage() treats the
+		// model as R1 (deepseek-reasoner id or the R1 format toggle), also treat it
+		// as preserving reasoning so the chain is fed back.
+		return {
+			id,
+			info: this.usesR1Format(id) ? { ...info, preserveReasoning: true } : info,
+			...params,
+		}
 	}
 
 	async completePrompt(prompt: string, options?: CompletePromptOptions): Promise<string> {

src/core/task/__tests__/reasoning-preservation.test.ts

@@ -512,4 +512,70 @@ describe("Task reasoning preservation", () => {
 			text: assistantText,
 		})
 	})
+
+	it("should keep plain text reasoning in buildCleanConversationHistory when preserveReasoning is true", async () => {
+		const task = new Task({
+			provider: mockProvider as ClineProvider,
+			apiConfiguration: mockApiConfiguration,
+			task: "Test task",
+			startTask: false,
+		})
+
+		const requestModelInfo: ModelInfo = {
+			contextWindow: 16000,
+			supportsPromptCache: true,
+			preserveReasoning: true,
+		}
+
+		await task["addToApiConversationHistory"](
+			{
+				role: "assistant",
+				content: [{ type: "text", text: "Reading the file." }],
+			},
+			"I should read the file first.",
+		)
+
+		const history = task["buildCleanConversationHistory"](task.apiConversationHistory, requestModelInfo)
+
+		expect(history).toHaveLength(1)
+		expect(history[0]).toMatchObject({
+			role: "assistant",
+			content: [
+				{ type: "reasoning", text: "I should read the file first.", summary: [] },
+				{ type: "text", text: "Reading the file." },
+			],
+		})
+	})
+
+	it("should strip plain text reasoning in buildCleanConversationHistory when preserveReasoning is not set", async () => {
+		const task = new Task({
+			provider: mockProvider as ClineProvider,
+			apiConfiguration: mockApiConfiguration,
+			task: "Test task",
+			startTask: false,
+		})
+
+		// preserveReasoning is undefined on this model info
+		const requestModelInfo: ModelInfo = {
+			contextWindow: 16000,
+			supportsPromptCache: true,
+		}
+
+		await task["addToApiConversationHistory"](
+			{
+				role: "assistant",
+				content: [{ type: "text", text: "Reading the file." }],
+			},
+			"I should read the file first.",
+		)
+
+		const history = task["buildCleanConversationHistory"](task.apiConversationHistory, requestModelInfo)
+
+		expect(history).toHaveLength(1)
+		expect(history[0]).toMatchObject({
+			role: "assistant",
+			content: "Reading the file.",
+		})
+		expect(JSON.stringify(history)).not.toContain("reasoning")
+	})
 })