sgl-project/sglang · error · ValueError

return_meta_info is not supported with streaming. Please set

Error message

return_meta_info is not supported with streaming. Please set stream=false when using return_meta_info=true.

What it means

return_meta_info=true is only supported for non-streaming /v1/chat/completions requests; meta info (finish details, usage internals, etc.) is attached to the complete GenerateOutput object. Streaming requests with this flag are rejected.

Source

Thrown at python/sglang/srt/entrypoints/openai/serving_chat.py:997

            )

        if reasoning_effort is not None:
            request.reasoning_effort = reasoning_effort

        if request.stream:
            if request.return_prompt_token_ids:
                raise ValueError(
                    "return_prompt_token_ids is not supported with streaming. "
                    "Please set stream=false when using return_prompt_token_ids=true."
                )
            if request.return_token_ids:
                raise ValueError(
                    "return_token_ids is not supported with streaming on "
                    "/v1/chat/completions. Please set stream=false when using "
                    "return_token_ids=true."
                )
            if request.return_meta_info:
                raise ValueError(
                    "return_meta_info is not supported with streaming. "
                    "Please set stream=false when using return_meta_info=true."
                )

        is_multimodal = self.tokenizer_manager.model_config.is_multimodal

        # Process messages and apply chat template
        processed_messages = self._process_messages(request, is_multimodal)
        # Build sampling parameters
        sampling_params = request.to_sampling_params(
            stop=processed_messages.stop,
            model_generation_config=self.default_sampling_params,
            tool_call_constraint=processed_messages.tool_call_constraint,
            renderer_handles_response_format=self.chat_encoding_spec == "kimi_k3",
        )

        # Handle single vs multiple requests
        if request.input_ids is not None:

View on GitHub (pinned to 0132848349)

Solutions

1. Set stream=false when you need meta_info
2. Or remove return_meta_info for streaming — finish_reason/usage still arrive via SSE events where supported

Example fix

// before
{"stream": true, "return_meta_info": true}
// after
{"stream": false, "return_meta_info": true}

if body.get("stream"):
    body["return_meta_info"] = False

def stream_compat_ok(b): return not (b.get('stream') and b.get('return_meta_info'))

When it happens

Trigger: POST /v1/chat/completions with {"stream": true, "return_meta_info": true}.

Common situations: Inspection/debug tooling built on the non-streaming API being switched to streaming; wanting per-chunk metadata like finish reason (which is already in SSE chunks).

Related errors

• return_prompt_token_ids is not supported with streaming. Ple
• return_token_ids is not supported with streaming on /v1/chat
• This use case is not supported. For OpenAI chat models, sgl.
• invalid reasoning effort: {effort!r}
• tool_choice 'required' or a named tool cannot be combined wi

AI-assisted analysis of sgl-project/sglang@0132848349 (2026-08-28). Data as JSON: /api/errors/4a434d78a78f32cf. Report an issue: GitHub.