diff --git a/vllm/entrypoints/openai/api_server.py b/vllm/entrypoints/openai/api_server.py
index a354ab0d5a05..5ae0e65d0093 100644
--- a/vllm/entrypoints/openai/api_server.py
+++ b/vllm/entrypoints/openai/api_server.py
@@ -124,7 +124,11 @@ async def create_completion(raw_request: Request):
 
     model_name = request.model
     request_id = f"cmpl-{random_uuid()}"
-    prompt = request.prompt
+    if isinstance(request.prompt, list):
+        assert len(request.prompt) == 1
+        prompt = request.prompt[0]
+    else:
+        prompt = request.prompt
     created_time = int(time.time())
     try:
         sampling_params = SamplingParams(
diff --git a/vllm/entrypoints/openai/protocol.py b/vllm/entrypoints/openai/protocol.py
index a6ef644d055c..01ec547ae45f 100644
--- a/vllm/entrypoints/openai/protocol.py
+++ b/vllm/entrypoints/openai/protocol.py
@@ -67,7 +67,7 @@ class ChatCompletionRequest(BaseModel):
 
 class CompletionRequest(BaseModel):
     model: str
-    prompt: str
+    prompt: Union[str, List[str]]
     suffix: Optional[str] = None
     max_tokens: Optional[int] = 16
     temperature: Optional[float] = 1.0