Comfy-Org/ComfyUI · error · ValueError

The provider rejected the audio track this model generated f

Error message

The provider rejected the audio track this model generated for the video (possible copyright match). The video itself was fine. Turn off generate_audio to get a silent video, or adjust the prompt and try again.

What it means

Raised by the Seedance 2 node after the BytePlus task poll fails and the provider error string contains 'OutputAudioSensitiveContentDetected.PolicyViolation'. The provider's content filter flagged the generated audio track (often a copyright/music match), even though the video frames themselves passed. The node re-raises as ValueError with a targeted remediation message instead of surfacing the raw API failure.

Source

Thrown at comfy_api_nodes/nodes_bytedance.py:2286

    task_id: str,
    model_id: str,
    resolution: str,
    has_video_input: bool,
) -> TaskStatusResponse:
    try:
        return await poll_op(
            cls,
            ApiEndpoint(path=f"{BYTEPLUS_SEEDANCE2_TASK_STATUS_ENDPOINT}/{task_id}"),
            response_model=TaskStatusResponse,
            status_extractor=lambda r: r.status,
            price_extractor=_seedance2_price_extractor(
                model_id, has_video_input=has_video_input, resolution=resolution
            ),
            poll_interval=9,
        )
    except Exception as exc:
        if _SEEDANCE_AUDIO_POLICY_CODE in str(exc):
            raise ValueError(
                "The provider rejected the audio track this model generated for the video "
                "(possible copyright match). The video itself was fine. Turn off generate_audio "
                "to get a silent video, or adjust the prompt and try again."
            ) from exc
        if _SEEDANCE_TASK_TYPE_CONSTRAINT_CODE in str(exc):
            raise ValueError(
                "Seedance read this prompt as editing the reference video, and an edit always "
                "takes its duration and aspect ratio from that video. Enable video_editing on "
                "this node and run again, or reword the prompt so it describes a new video "
                "rather than a change to the reference one."
            ) from exc
        raise


def _seedance2_price_badge(with_reference_videos: bool) -> IO.PriceBadge:
    widgets = ["model", "model.resolution", "model.ratio", "model.duration"]
    if with_reference_videos:
        widgets.append("model.video_editing")

View on GitHub (pinned to 1c6d8d45b3)

Solutions

Rerun with generate_audio off to get a silent video (the video content itself was accepted).
Reword the prompt to remove song names, artist names, or recognizable melodies and retry with audio on.
If the prompt is clearly original content, generate audio separately (e.g. with the Seed-OSS audio node) and mux it locally.

Defensive patterns

Strategy: retry

Try / catch

try:
    out = await node_api_call(...)
except ValueError as e:
    if "possible copyright match" in str(e):
        rerun_with(generate_audio=False)  # silent video path
    else:
        raise

Prevention

Avoid artist names, song titles, and recognizable melodies in prompts when generate_audio is on.
Wrap Seedance audio-enabled runs in a workflow branch that can retry with generate_audio disabled.
Generate audio in a separate node so a policy hit only costs the audio stage.

When it happens

Trigger: Calling the Seedance 2 generate node with generate_audio enabled and a prompt whose generated audio resembles copyrighted music/lyrics; poll_op raises, _SEEDANCE_AUDIO_POLICY_CODE ("OutputAudioSensitiveContentDetected.PolicyViolation") is found in str(exc), and this ValueError replaces it.

Common situations: Prompts naming real songs, artists, or humming famous melodies; sound-effect prompts that trigger the audio fingerprinter; happens intermittently on retry because detection is heuristic.

Related errors

AI-assisted analysis of Comfy-Org/ComfyUI@1c6d8d45b3 (2026-08-14). Data as JSON: /api/errors/07aeeb6620beb097. Report an issue: GitHub.