diff --git a/lmdeploy/serve/openai/api_server.py b/lmdeploy/serve/openai/api_server.py
index 28096af9b..e1af990a5 100644
--- a/lmdeploy/serve/openai/api_server.py
+++ b/lmdeploy/serve/openai/api_server.py
@@ -60,7 +60,7 @@ def create_error_response(status: HTTPStatus, message: str):
     return JSONResponse(
         ErrorResponse(message=message,
                       type='invalid_request_error',
-                      code=status.value).dict())
+                      code=status.value).model_dump())
 
 
 async def check_request(request) -> Optional[JSONResponse]:
@@ -152,7 +152,7 @@ def create_stream_response_json(
             model=model_name,
             choices=[choice_data],
         )
-        response_json = response.json(ensure_ascii=False)
+        response_json = response.model_dump_json()
 
         return response_json
 
@@ -167,7 +167,7 @@ async def completion_stream_generator() -> AsyncGenerator[str, None]:
             chunk = ChatCompletionStreamResponse(id=request_id,
                                                  choices=[choice_data],
                                                  model=model_name)
-            data = chunk.json(exclude_unset=True, ensure_ascii=False)
+            data = chunk.model_dump_json(exclude_unset=True)
             yield f'data: {data}\n\n'
 
         async for res in result_generator:
diff --git a/lmdeploy/serve/openai/protocol.py b/lmdeploy/serve/openai/protocol.py
index 8d5b38757..8f2919a1a 100644
--- a/lmdeploy/serve/openai/protocol.py
+++ b/lmdeploy/serve/openai/protocol.py
@@ -84,7 +84,7 @@ class ChatCompletionResponseChoice(BaseModel):
     """Chat completion response choices."""
     index: int
     message: ChatMessage
-    finish_reason: Optional[Literal['stop', 'length']]
+    finish_reason: Optional[Literal['stop', 'length']] = None
 
 
 class ChatCompletionResponse(BaseModel):
@@ -107,7 +107,7 @@ class ChatCompletionResponseStreamChoice(BaseModel):
     """Chat completion response stream choice."""
     index: int
     delta: DeltaMessage
-    finish_reason: Optional[Literal['stop', 'length']]
+    finish_reason: Optional[Literal['stop', 'length']] = None
 
 
 class ChatCompletionStreamResponse(BaseModel):
@@ -142,7 +142,7 @@ class CompletionResponseChoice(BaseModel):
     index: int
     text: str
     logprobs: Optional[int] = None
-    finish_reason: Optional[Literal['stop', 'length']]
+    finish_reason: Optional[Literal['stop', 'length']] = None
 
 
 class CompletionResponse(BaseModel):