diff --git a/basic_demo/openai_api_server.py b/basic_demo/openai_api_server.py
index 99a159f..7b78b26 100644
--- a/basic_demo/openai_api_server.py
+++ b/basic_demo/openai_api_server.py
@@ -446,9 +446,10 @@ async def predict_stream(model_id, gen_params):
     response_id = generate_id('chatcmpl-', 29)
     system_fingerprint = generate_id('fp_', 9)
     tools = {tool['function']['name'] for tool in gen_params['tools']} if gen_params['tools'] else {}
+    delta_text = ""
     async for new_response in generate_stream_glm4(gen_params):
         decoded_unicode = new_response["text"]
-        delta_text = decoded_unicode[len(output):]
+        delta_text += decoded_unicode[len(output):]
         output = decoded_unicode
         lines = output.strip().split("\n")
 
@@ -461,6 +462,7 @@ async def predict_stream(model_id, gen_params):
             if first_line in tools:
                 is_function_call = True
                 function_name = first_line
+                delta_text = lines[1]
 
         # 工具调用返回
         if is_function_call:
@@ -496,6 +498,7 @@ async def predict_stream(model_id, gen_params):
                 has_send_first_chunk = True
 
             function_call = {"name": None, "arguments": delta_text}
+            delta_text = ""
             tool_call = ChatCompletionMessageToolCall(
                 index=0,
                 id=None,
@@ -557,6 +560,7 @@ async def predict_stream(model_id, gen_params):
                 role="assistant",
                 function_call=None,
             )
+            delta_text = ""
             choice_data = ChatCompletionResponseStreamChoice(
                 index=0,
                 delta=message,
@@ -592,8 +596,51 @@ async def predict_stream(model_id, gen_params):
             object="chat.completion.chunk",
             usage=None
         ).model_dump_json(exclude_unset=True)
-    yield '[DONE]'
-
+    elif delta_text != "":
+        message = DeltaMessage(
+            content="",
+            role="assistant",
+            function_call=None,
+        )
+        choice_data = ChatCompletionResponseStreamChoice(
+            index=0,
+            delta=message,
+            finish_reason=None
+        )
+        chunk = ChatCompletionResponse(
+            model=model_id,
+            id=response_id,
+            choices=[choice_data],
+            created=created_time,
+            system_fingerprint=system_fingerprint,
+            object="chat.completion.chunk"
+        )
+        yield chunk.model_dump_json(exclude_unset=True)
+    
+        finish_reason = 'stop'
+        message = DeltaMessage(
+            content=delta_text,
+            role="assistant",
+            function_call=None,
+        )
+        delta_text = ""
+        choice_data = ChatCompletionResponseStreamChoice(
+            index=0,
+            delta=message,
+            finish_reason=finish_reason
+        )
+        chunk = ChatCompletionResponse(
+            model=model_id,
+            id=response_id,
+            choices=[choice_data],
+            created=created_time,
+            system_fingerprint=system_fingerprint,
+            object="chat.completion.chunk"
+        )
+        yield chunk.model_dump_json(exclude_unset=True)
+        yield '[DONE]'
+    else:
+        yield '[DONE]'
 
 async def parse_output_text(model_id: str, value: str, function_call: ChoiceDeltaToolCallFunction = None):
     delta = DeltaMessage(role="assistant", content=value)