diff --git a/vllm/entrypoints/openai/serving_chat.py b/vllm/entrypoints/openai/serving_chat.py
index 1ea18c1d3300e..1dee798858a16 100644
--- a/vllm/entrypoints/openai/serving_chat.py
+++ b/vllm/entrypoints/openai/serving_chat.py
@@ -146,9 +146,8 @@ async def create_chat_completion(
             sampling_params = request.to_sampling_params(
                 tokenizer, guided_decode_logits_processor)
             if sampling_params.max_tokens is None:
-                sampling_params.max_tokens = (
-                    self.max_model_len -
-                    len(prompt_inputs["prompt_token_ids"]))
+                sampling_params.max_tokens = \
+                    self.max_model_len - len(prompt_inputs["prompt_token_ids"])
 
             self._log_inputs(request_id,
                              prompt_inputs,
diff --git a/vllm/entrypoints/openai/serving_completion.py b/vllm/entrypoints/openai/serving_completion.py
index 226cb5e04c5dd..51de2f1826554 100644
--- a/vllm/entrypoints/openai/serving_completion.py
+++ b/vllm/entrypoints/openai/serving_completion.py
@@ -108,9 +108,8 @@ async def create_completion(self, request: CompletionRequest,
                 sampling_params = request.to_sampling_params(
                     tokenizer, guided_decode_logits_processor)
                 if sampling_params.max_tokens is None:
-                    sampling_params.max_tokens = (
-                        self.max_model_len -
-                        len(prompt_inputs["prompt_token_ids"]))
+                    sampling_params.max_tokens = self.max_model_len - \
+                        len(prompt_inputs["prompt_token_ids"])
 
                 request_id_item = f"{request_id}-{i}"