calesthio/OpenMontage · error · RuntimeError

Seedream task {status}: {error_msg}

Error message

Seedream task {status}: {error_msg}

What it means

Raised when the polled Seedream queue task reports status 'FAILED' or 'CANCELLED'. The message embeds the status and the API-supplied 'error' field (defaulting to 'Unknown error' when absent). This is the server-side generation actually failing — moderation rejection, invalid parameters, or upstream model error.

Source

Thrown at tools/graphics/seedream_image.py:216

                f"https://queue.fal.run/bytedance/seedream/requests/"
                f"{request_id}/status"
            )
            elapsed = 0.0
            while elapsed < 300:
                status_resp = requests.get(
                    status_url,
                    headers=headers,
                    timeout=30,
                )
                status_resp.raise_for_status()
                status_data = status_resp.json()
                status = status_data.get("status")

                if status == "COMPLETED":
                    break
                elif status in ("FAILED", "CANCELLED"):
                    error_msg = status_data.get("error", "Unknown error")
                    raise RuntimeError(f"Seedream task {status}: {error_msg}")

                time.sleep(10)
                elapsed += 10

            if elapsed >= 300:
                raise RuntimeError(
                    f"Seedream task timed out after {300}s"
                )

            result_resp = requests.get(
                f"https://queue.fal.run/bytedance/seedream/requests/"
                f"{request_id}",
                headers=headers,
                timeout=30,
            )
            result_resp.raise_for_status()
            result_data = result_resp.json()

View on GitHub (pinned to 95e1c3d0ab)

Solutions

1. Read the embedded error message — moderation vs internal error have different fixes
2. For moderation: soften or rephrase the prompt, or swap the reference image
3. For transient upstream errors: retry the same inputs once or twice
4. For 'CANCELLED': check whether another process/dashboard cancelled the request

null

try:
    paths = seedream_run(inputs)
except RuntimeError as e:
    msg = str(e)
    if "FAILED" in msg or "CANCELLED" in msg:
        if "moderation" in msg.lower():
            inputs = soften_prompt(inputs)  # rephrase and resubmit
        else:
            retry_with_backoff(seedream_run, inputs, attempts=2)  # transient upstream
    else:
        raise

When it happens

Trigger: Prompt or reference image tripped content moderation; image reference URL unreachable by the backend; internal Seedream model error during render; task manually cancelled in the fal dashboard.

Common situations: borderline NSFW or celebrity-likeness prompts; expired/signed reference URLs passed after their TTL; occasional upstream capacity errors that look like permanent failures.

Related errors

• Seedream submit succeeded but did not return request_id
• Seedream task timed out after {300}s
• Seedream completed but no images returned
• duration_seconds is required for cost estimation
• model_id must be one of: {choices}

AI-assisted analysis of calesthio/OpenMontage@95e1c3d0ab (2026-08-15). Data as JSON: /api/errors/589bdde7d0ba1532. Report an issue: GitHub.