diff --git a/vllm/entrypoints/openai/api_server.py b/vllm/entrypoints/openai/api_server.py
index be5f4190e633f..8dc2e58e4709d 100644
--- a/vllm/entrypoints/openai/api_server.py
+++ b/vllm/entrypoints/openai/api_server.py
@@ -272,6 +272,8 @@ async def create_chat_completion(request: ChatCompletionRequest,
             top_k=request.top_k,
             ignore_eos=request.ignore_eos,
             use_beam_search=request.use_beam_search,
+            logprobs=request.logprobs,
+            prompt_logprobs=request.logprobs if request.echo else None,
             skip_special_tokens=request.skip_special_tokens,
             spaces_between_special_tokens=spaces_between_special_tokens,
         )
@@ -292,7 +294,11 @@ async def completion_stream_generator() -> AsyncGenerator[str, None]:
         role = get_role()
         for i in range(request.n):
             choice_data = ChatCompletionResponseStreamChoice(
-                index=i, delta=DeltaMessage(role=role), finish_reason=None)
+                index=i,
+                delta=DeltaMessage(role=role),
+                logprobs=(LogProbs()
+                          if request.logprobs is not None else None),
+                finish_reason=None)
             chunk = ChatCompletionStreamResponse(id=request_id,
                                                  object=chunk_object_type,
                                                  created=created_time,
@@ -314,6 +320,8 @@ async def completion_stream_generator() -> AsyncGenerator[str, None]:
                     choice_data = ChatCompletionResponseStreamChoice(
                         index=i,
                         delta=DeltaMessage(content=last_msg_content),
+                        logprobs=(LogProbs()
+                                  if request.logprobs is not None else None),
                         finish_reason=None)
                     chunk = ChatCompletionStreamResponse(
                         id=request_id,
@@ -339,11 +347,24 @@ async def completion_stream_generator() -> AsyncGenerator[str, None]:
                 if output.finish_reason is None:
                     # Send token-by-token response for each request.n
                     delta_text = output.text[len(previous_texts[i]):]
+                    if request.logprobs is not None:
+                        token_ids = output.token_ids[previous_num_tokens[i]:]
+                        top_logprobs = output.logprobs[previous_num_tokens[i]:]
+                        offsets = len(previous_texts[i])
+                        logprobs = create_logprobs(
+                            token_ids=token_ids,
+                            top_logprobs=top_logprobs,
+                            num_output_top_logprobs=request.logprobs,
+                            initial_text_offset=offsets,
+                        )
+                    else:
+                        logprobs = None
                     previous_texts[i] = output.text
                     previous_num_tokens[i] = len(output.token_ids)
                     choice_data = ChatCompletionResponseStreamChoice(
                         index=i,
                         delta=DeltaMessage(content=delta_text),
+                        logprobs=logprobs,
                         finish_reason=None)
                     chunk = ChatCompletionStreamResponse(
                         id=request_id,
@@ -362,7 +383,11 @@ async def completion_stream_generator() -> AsyncGenerator[str, None]:
                         total_tokens=prompt_tokens + previous_num_tokens[i],
                     )
                     choice_data = ChatCompletionResponseStreamChoice(
-                        index=i, delta=[], finish_reason=output.finish_reason)
+                        index=i,
+                        delta=[],
+                        logprobs=(LogProbs()
+                                  if request.logprobs is not None else None),
+                        finish_reason=output.finish_reason)
                     chunk = ChatCompletionStreamResponse(
                         id=request_id,
                         object=chunk_object_type,
@@ -391,11 +416,27 @@ async def completion_full_generator():
         assert final_res is not None
 
         choices = []
+        prompt_token_ids = final_res.prompt_token_ids
+        prompt_logprobs = final_res.prompt_logprobs
         role = get_role()
         for output in final_res.outputs:
+            if request.logprobs is not None:
+                token_ids = output.token_ids
+                top_logprobs = output.logprobs
+                if request.echo:
+                    token_ids = prompt_token_ids + token_ids
+                    top_logprobs = prompt_logprobs + top_logprobs
+                logprobs = create_logprobs(
+                    token_ids=token_ids,
+                    top_logprobs=top_logprobs,
+                    num_output_top_logprobs=request.logprobs,
+                )
+            else:
+                logprobs = None
             choice_data = ChatCompletionResponseChoice(
                 index=output.index,
                 message=ChatMessage(role=role, content=output.text),
+                logprobs=logprobs,
                 finish_reason=output.finish_reason,
             )
             choices.append(choice_data)
diff --git a/vllm/entrypoints/openai/protocol.py b/vllm/entrypoints/openai/protocol.py
index 7a86a19c4bf80..e08cc6d144f93 100644
--- a/vllm/entrypoints/openai/protocol.py
+++ b/vllm/entrypoints/openai/protocol.py
@@ -61,6 +61,7 @@ class ChatCompletionRequest(BaseModel):
     max_tokens: Optional[int] = None
     stop: Optional[Union[str, List[str]]] = Field(default_factory=list)
     stream: Optional[bool] = False
+    logprobs: Optional[int] = None
     presence_penalty: Optional[float] = 0.0
     frequency_penalty: Optional[float] = 0.0
     logit_bias: Optional[Dict[str, float]] = None
@@ -155,6 +156,7 @@ class ChatMessage(BaseModel):
 class ChatCompletionResponseChoice(BaseModel):
     index: int
     message: ChatMessage
+    logprobs: Optional[LogProbs] = None
     finish_reason: Optional[Literal["stop", "length"]] = None
 
 
@@ -175,6 +177,7 @@ class DeltaMessage(BaseModel):
 class ChatCompletionResponseStreamChoice(BaseModel):
     index: int
     delta: DeltaMessage
+    logprobs: Optional[LogProbs] = None
     finish_reason: Optional[Literal["stop", "length"]] = None