update

app/core/runner/llm_backend.py

@@ -35,7 +35,7 @@ class LLMBackend:
             "messages": messages,
             "model": model,
             "stream": stream,
-            "parallel_tool_calls": False,
+            "parallel_tool_calls": parallel_tool_calls,
         }
         if extra_body:
             model_params = extra_body.get("model_params")
@@ -57,8 +57,6 @@ class LLMBackend:
                     chat_params["stream_options"] = {
                         "include_usage": bool(stream_options["include_usage"])
                     }
-        # if parallel_tool_calls:
-        chat_params["parallel_tool_calls"] = parallel_tool_calls
         if audio:
             chat_params["audio"] = audio
         if modalities:

app/core/runner/thread_runner.py

@@ -159,9 +159,9 @@ class ThreadRunner:
             temperature=run.temperature,
             top_p=run.top_p,
             response_format=run.response_format,
-            # parallel_tool_calls=run.parallel_tool_calls,
-            # audio=run.audio,
-            # modalities=run.modalities,
+            parallel_tool_calls=run.parallel_tool_calls,
+            audio=run.audio,
+            modalities=run.modalities,
         )
 
         # create message callback