Update app.py

app.py

@@ -159,11 +159,13 @@ def generate(
 
         final_response = "".join(outputs)
 
+        # Yield text response first
+        yield final_response  
+
         if is_tts and voice:
             output_file = asyncio.run(text_to_speech(final_response, voice))
-            yield gr.Audio(output_file, autoplay=True)  # Return audio response as a separate component
-        else:
-            yield final_response  # Return text response
+            # Return playable audio separately
+            yield gr.Audio(output_file, autoplay=True)
 
 demo = gr.ChatInterface(
     fn=generate,