fix: omit Ollama think for non-reasoning models

Preserve native Ollama thinking controls for supported models and explicit think=false, but avoid sending truthy think payloads for models marked reasoning=false.\n\nCo-authored-by: 吴杨帆 <85487201+leno23@users.noreply.github.com>

CHANGELOG.md

@@ -36,6 +36,7 @@ Docs: https://docs.openclaw.ai
 - Gateway/WebSocket: log expected startup `1013 gateway starting` retry closes at debug instead of warn while preserving WARN for unexpected pre-connect failures. Fixes #76361. (#82457) Thanks @IWhatsskill.
 - Providers/Xiaomi: strip synthetic empty array `items` from MiMo tool schemas while preserving typed array items, avoiding strict OpenAI-compatible schema rejection.
 - Telegram: send the transcript-backed full final answer after progress-mode tool drafts when the dispatcher final payload is an ellipsis-truncated snapshot. Fixes #82409. Thanks @PashaGanson.
+- Providers/Ollama: omit truthy native `think` payloads for models marked non-reasoning while preserving supported thinking models and explicit `think: false`. (#82445) Thanks @leno23.
 - CLI/context engines: bootstrap and finalize non-legacy context engines for CLI turns while preserving transcript snapshots and deferred maintenance ownership. (#81869) Thanks @sahilsatralkar.
 - Telegram: persist polling updates through restart replay so queued same-topic messages resume in order instead of losing context after a gateway restart. (#82256) Thanks @VACInc.
 - Gateway/Gmail: abort in-flight Gmail watcher startup and hot-reload restarts before shutdown so reloads cannot spawn `gog serve` after the Gateway is closing. Thanks @frankekn.

extensions/ollama/src/stream-runtime.test.ts

@@ -292,6 +292,113 @@ describe("createConfiguredOllamaCompatStreamWrapper", () => {
     );
   });
 
+  it("does not forward truthy configured native Ollama thinking for non-reasoning models", async () => {
+    await withMockNdjsonFetch(
+      [
+        '{"model":"m","created_at":"t","message":{"role":"assistant","content":"ok"},"done":false}',
+        '{"model":"m","created_at":"t","message":{"role":"assistant","content":""},"done":true,"prompt_eval_count":1,"eval_count":1}',
+      ],
+      async (fetchMock) => {
+        const baseStreamFn = createOllamaStreamFn("http://ollama-host:11434");
+        const model = {
+          api: "ollama",
+          provider: "ollama",
+          id: "llama3.2:latest",
+          contextWindow: 8192,
+          reasoning: false,
+          params: { thinking: "medium" },
+        };
+
+        const wrapped = createConfiguredOllamaCompatStreamWrapper({
+          provider: "ollama",
+          modelId: "llama3.2:latest",
+          model,
+          streamFn: baseStreamFn,
+          thinkingLevel: "off",
+        } as never);
+        if (!wrapped) {
+          throw new Error("Expected wrapped Ollama stream function");
+        }
+
+        const stream = await Promise.resolve(
+          wrapped(
+            model as never,
+            {
+              messages: [{ role: "user", content: "hello" }],
+            } as never,
+            {} as never,
+          ),
+        );
+
+        await collectStreamEvents(stream);
+
+        const requestInit = getGuardedFetchCall(fetchMock).init ?? {};
+        if (typeof requestInit.body !== "string") {
+          throw new Error("Expected string request body");
+        }
+        const requestBody = JSON.parse(requestInit.body) as {
+          think?: string;
+          options?: { think?: string };
+        };
+        expect(requestBody.think).toBeUndefined();
+        expect(requestBody.options?.think).toBeUndefined();
+      },
+    );
+  });
+
+  it("does not forward runtime native Ollama thinking for non-reasoning models", async () => {
+    await withMockNdjsonFetch(
+      [
+        '{"model":"m","created_at":"t","message":{"role":"assistant","content":"ok"},"done":false}',
+        '{"model":"m","created_at":"t","message":{"role":"assistant","content":""},"done":true,"prompt_eval_count":1,"eval_count":1}',
+      ],
+      async (fetchMock) => {
+        const baseStreamFn = createOllamaStreamFn("http://ollama-host:11434");
+        const model = {
+          api: "ollama",
+          provider: "ollama",
+          id: "llama3.2:latest",
+          contextWindow: 8192,
+          reasoning: false,
+        };
+
+        const wrapped = createConfiguredOllamaCompatStreamWrapper({
+          provider: "ollama",
+          modelId: "llama3.2:latest",
+          model,
+          streamFn: baseStreamFn,
+          thinkingLevel: "low",
+        } as never);
+        if (!wrapped) {
+          throw new Error("Expected wrapped Ollama stream function");
+        }
+
+        const stream = await Promise.resolve(
+          wrapped(
+            model as never,
+            {
+              messages: [{ role: "user", content: "hello" }],
+            } as never,
+            {} as never,
+          ),
+        );
+
+        await collectStreamEvents(stream);
+
+        const requestInit = getGuardedFetchCall(fetchMock).init ?? {};
+        if (typeof requestInit.body !== "string") {
+          throw new Error("Expected string request body");
+        }
+        const requestBody = JSON.parse(requestInit.body) as {
+          think?: string;
+          options?: { think?: string };
+        };
+        expect(requestBody.think).toBeUndefined();
+        expect(requestBody.options?.think).toBeUndefined();
+      },
+    );
+  });
+
   it("forwards the native think effort on native Ollama chat requests when thinking is enabled", async () => {
     await withMockNdjsonFetch(
       [

extensions/ollama/src/stream.ts

@@ -278,6 +278,15 @@ function resolveOllamaThinkParamValue(
   return undefined;
 }
 
+function shouldForwardNativeOllamaThink(
+  model: ProviderRuntimeModel | undefined,
+  think: OllamaThinkValue,
+): boolean {
+  // Ollama accepts top-level `think` as the native chat contract, but rejects
+  // truthy values for models known not to expose thinking support.
+  return think === false || model?.reasoning !== false;
+}
+
 function resolveOllamaConfiguredNumCtx(model: ProviderRuntimeModel): number | undefined {
   const raw = model.params?.num_ctx;
   if (typeof raw !== "number" || !Number.isFinite(raw) || raw <= 0) {
@@ -341,7 +350,7 @@ function resolveOllamaTopLevelParams(
     }
   }
   const think = resolveOllamaThinkParamValue(params);
-  if (think !== undefined) {
+  if (think !== undefined && shouldForwardNativeOllamaThink(model, think)) {
     requestParams.think = think;
   }
   return Object.keys(requestParams).length > 0 ? requestParams : undefined;
@@ -390,7 +399,7 @@ export function createConfiguredOllamaCompatStreamWrapper(
     runtimeThinkValue === false && configuredThinkValue !== undefined
       ? undefined
       : runtimeThinkValue;
-  if (ollamaThinkValue !== undefined) {
+  if (ollamaThinkValue !== undefined && shouldForwardNativeOllamaThink(model, ollamaThinkValue)) {
     streamFn = createOllamaThinkingWrapper(streamFn, ollamaThinkValue);
   }