Comfy-Org/ComfyUI · error · RuntimeError

ByteDance request failed. Code: {response.error['code']}, me

Error message

ByteDance request failed. Code: {response.error['code']}, message: {response.error['message']}

What it means

ByteDance image tasks return an error object inside an otherwise-HTTP-200 response. get_image_url_from_response checks response.error and re-raises it as RuntimeError with the API's code and message, after logging it at info level.

Source

Thrown at comfy_api_nodes/nodes_bytedance.py:427

def _seedance2_price_extractor(model_id: str, has_video_input: bool, resolution: str):
    """Returns a price_extractor closure for Seedance 2.0 poll_op."""
    rate = seedance2_price_per_1k_tokens(model_id, has_video_input, resolution)
    if rate is None:
        return None

    def extractor(response: TaskStatusResponse) -> float | None:
        if response.usage is None:
            return None
        return response.usage.total_tokens * 1.43 * rate / 1_000.0

    return extractor


def get_image_url_from_response(response: ImageTaskCreationResponse) -> str:
    if response.error:
        error_msg = f"ByteDance request failed. Code: {response.error['code']}, message: {response.error['message']}"
        logging.info(error_msg)
        raise RuntimeError(error_msg)
    logging.info("ByteDance task succeeded, image URL: %s", response.data[0]["url"])
    return response.data[0]["url"]


class ByteDanceImageNode(IO.ComfyNode):

    @classmethod
    def define_schema(cls):
        return IO.Schema(
            node_id="ByteDanceImageNode",
            display_name="ByteDance Image",
            category="partner/image/ByteDance",
            description="Generate images using ByteDance models via api based on prompt",
            inputs=[
                IO.Combo.Input("model", options=["seedream-3-0-t2i-250415"]),
                IO.String.Input(
                    "prompt",
                    multiline=True,

View on GitHub (pinned to 1c6d8d45b3)

Solutions

Read the code/message pair: content-policy rejections need a prompt change; quota errors need billing/credits.
Fix the parameter the code points at (size, seed range, model id).
Transient codes can be handled by re-queueing the workflow once.

Defensive patterns

Strategy: try-catch

Try / catch

try:
    url = get_image_url_from_response(response)
except RuntimeError as e:
    code = parse_code(str(e))
    if code in TRANSIENT_CODES:
        response = await recreate_task(...)  # single retry
    else:
        raise  # moderation/quota errors need human action

Prevention

Check response.error on every ByteDance task response instead of assuming 200 means success.
Keep prompts within content policy and monitor account credits before long batch runs.

When it happens

Trigger: Any ByteDanceImageNode task whose ImageTaskCreationResponse carries a non-null error - invalid prompt (moderation), quota exhausted, invalid parameters, or content policy rejection.

Common situations: Prompt trips content moderation; account out of credits; model name or size string rejected by the API; transient upstream failure during task creation.

Related errors

AI-assisted analysis of Comfy-Org/ComfyUI@1c6d8d45b3 (2026-08-14). Data as JSON: /api/errors/5750acebe36689de. Report an issue: GitHub.