Support protocol thread configuration

tests/test_direct_protocol.py

@@ -198,6 +198,23 @@ class OllamaTests(unittest.TestCase):
         self.assertFalse(payload["stream"])
         self.assertEqual(result.raw_response, raw)
 
+    def test_request_omits_num_thread_when_backend_selection_is_requested(self) -> None:
+        response = Mock()
+        response.raise_for_status.return_value = None
+        response.json.return_value = {"response": "# Meeting Protocol"}
+        with patch.object(ollama.requests, "post", return_value=response) as post:
+            ollama.generate_once(
+                "http://localhost:11434",
+                "qwen3.8:27b",
+                "prompt",
+                timeout=30,
+                num_ctx=32768,
+                num_predict=8192,
+            )
+
+        payload = post.call_args.kwargs["json"]
+        self.assertNotIn("num_thread", payload["options"])
+
     def test_malformed_response_failure_without_retry(self) -> None:
         response = Mock()
         response.raise_for_status.return_value = None