diff --git a/vllm/entrypoints/openai/serving_chat.py b/vllm/entrypoints/openai/serving_chat.py
index 79eac184a212..2a0d4cd74a28 100644
--- a/vllm/entrypoints/openai/serving_chat.py
+++ b/vllm/entrypoints/openai/serving_chat.py
@@ -873,7 +873,7 @@ class OpenAIServingChat(OpenAIServing):
                             total_tokens=num_prompt_tokens + completion_tokens,
                         )
 
-                    data = chunk.model_dump_json(exclude_none=True)
+                    data = chunk.model_dump_json(exclude_unset=True)
                     yield f"data: {data}\n\n"
 
             # once the final token is handled, if stream_options.include_usage