sgl-project/sglang · error · RuntimeError

No image data in response

Error message

No image data in response

What it means

Raised in the ComfyUI node when the server response contains no image payload (empty images array or missing b64 data). Typically a failed/blocked generation upstream; inspect the job status before reading image data.

Solutions

1. Log the raw response inside the node to inspect its shape.
2. If URLs are returned, fetch them instead of expecting b64_json.
3. Check server logs for why generation yielded no image data.
4. Retry — occasional transient empty payloads indicate server overload.

Example fix

// before
if not response["data"] or not response["data"][0]["b64_json"]:
    raise RuntimeError("No image data in response")
// after
entry = (response.get("data") or [{}])[0]
if not entry.get("b64_json"):
    raise RuntimeError(f"No image data in response: {response}")

entry = (response.get('data') or [{}])[0]
assert entry.get('b64_json'), f'unexpected payload: {response}'

def usable_image_response(r: dict) -> bool:
    data = r.get('data') or []
    return bool(data) and bool(data[0].get('b64_json'))

if not usable_image_response(response):
    raise RuntimeError(f'No image data in response: {response}')

When it happens

Trigger: Server returns 200 with empty data, or data[0] lacking b64_json (e.g. URL-mode response or silent per-image failure).

Common situations: Server bug or partial failure; response_format mismatch; requesting more images than the server produced.

Related errors

• Failed to generate image
• Prompt cannot be empty
• (empty data payload)
• (missing ;base64 marker)
• data URI must use ;base64 encoding

AI-assisted analysis of sgl-project/sglang@0132848349 (2026-08-28). Data as JSON: /api/errors/1b957d46d65141f2. Report an issue: GitHub.

Appendix: source

Thrown at python/sglang/multimodal_gen/apps/ComfyUI_SGLDiffusion/nodes.py:390

            request_params["enable_teacache"] = True
        if image is not None:
            # If the image is empty, use the size of the image to generate the image
            if is_empty_image(image):
                width, height = image.shape[2], image.shape[1]
                size = f"{width}x{height}"
                request_params["size"] = size
            else:
                request_params["image_path"] = get_image_path(image)

        # Call API
        try:
            response = sgld_client.generate_image(**request_params)
        except Exception as e:
            raise RuntimeError(f"Failed to generate image: {str(e)}")

        # Decode base64 image
        if not response["data"] or not response["data"][0]["b64_json"]:
            raise RuntimeError("No image data in response")
        image_data = response["data"][0]["b64_json"]
        image = convert_b64_to_tensor_image(image_data)

        return (image,)


class SGLDiffusionGenerateVideo:
    """Node to generate videos using SGLang Diffusion."""

    @classmethod
    def INPUT_TYPES(cls):
        return {
            "required": {
                "sgld_client": ("SGLD_CLIENT",),
                "positive_prompt": (
                    "STRING",
                    {
                        "default": "",

View on GitHub (pinned to 0132848349)