Errors

HTTP status codes and error codes returned by the Seedance 3.0 API, and when to retry.

Errors use standard HTTP status codes and a JSON body:

{
  "error": {
    "type": "insufficient_credits",
    "code": "INSUFFICIENT_CREDITS",
    "message": "Not enough credits. Top up at /pricing."
  }
}

code is present when the failure came from the generation pipeline. param names the offending field on validation errors.

Status codes

Statuserror.typeerror.codeWhat to do
400invalid_request_errorINVALID_INPUT, CONTENT_POLICYFix the request or prompt. Don't retry unchanged
401authentication_errorCheck the API key
402insufficient_creditsINSUFFICIENT_CREDITSTop up credits
403permission_errorThe account is disabled
404not_found_errorUnknown model id or generation id
429rate_limit_errorRATE_LIMIT, TOO_MANY_IN_FLIGHTBack off and retry; wait for running jobs to finish
502api_errorGENERATION_FAILEDProvider failed. Retry once
503api_errorSERVICE_UNAVAILABLERetry with exponential backoff
504api_errorTIMEOUTRetry with exponential backoff

Failed generations

A generation can also fail after it was created. In that case GET /generations/{id} returns 200 with status: "failed" and an error object using the same codes:

{
  "id": "0b6f4c1e-8f1a-4d0e-9a51-1d3c2a7e9b10",
  "object": "generation",
  "status": "failed",
  "outputs": [],
  "error": {
    "code": "CONTENT_POLICY",
    "message": "The prompt or input media was rejected by content moderation."
  }
}

Credits for failed generations are refunded automatically, except when the finished output is blocked by moderation or the job times out at the provider.

Limits

  • Requests to POST /generations from the same client are limited to roughly one per second.
  • Each account can have a limited number of generations running at once. Extra requests return 429 with TOO_MANY_IN_FLIGHT until one finishes.