Update app.py

app.py

@@ -109,11 +109,13 @@ def stream_response(message: Dict[str, Any], history: List[Dict[str, Any]], mode
         yield f"Failed to get response: {e}"
 
 def build_demo() -> gr.Blocks:
+    chatbot = gr.Chatbot(type="messages", allow_tags=["think"])
     textbox = gr.MultimodalTextbox(
         show_label=False,
         placeholder="Enter text, or upload one or more images...",
         file_types=["image"],
         file_count="multiple",
+        max_plain_text_length=32768
     )
     model_selector = gr.Dropdown(
         label="Model",
@@ -127,6 +129,7 @@ def build_demo() -> gr.Blocks:
         fn=stream_response,
         type="messages",
         multimodal=True,
+        chatbot=chatbot,
         textbox=textbox,
         title="Qianfan-VL: Domain-Enhanced Vision-Language Models",
         description="""**Qianfan-VL** is a series of general-purpose multimodal large language models enhanced for enterprise-level multimodal applications. The models offer deep optimization for high-frequency scenarios in industrial deployment while maintaining strong general capabilities.
@@ -140,4 +143,4 @@ def main():
     build_demo().launch()
 
 if __name__ == "__main__":
-    main()
+    main()