BerriAI/litellm · error · BlackForestLabsError

Request failed: {e}

Error message

Request failed: {e}

What it means

This wraps any non-HTTP exception from the synchronous initial POST that submits the image-generation job to BFL (sync_client.post in image_generation/handler.py). Network errors, DNS failures, TLS problems, and timeouts surface here as a 500 BlackForestLabsError with the underlying exception text. It fires before polling begins, so no job was created.

Source

Thrown at litellm/llms/black_forest_labs/image_generation/handler.py:157

            input=prompt,
            api_key="",
            additional_args={
                "complete_input_dict": data,
                "api_base": complete_url,
                "headers": headers,
            },
        )

        # Make initial request
        try:
            response: Final = sync_client.post(
                url=complete_url,
                headers=headers,
                json=data,
                timeout=timeout,
            )
        except Exception as e:
            raise BlackForestLabsError(
                status_code=500,
                message=f"Request failed: {e}",
            )

        # Poll for result
        final_response: Final = self._poll_for_result_sync(
            initial_response=response,
            headers=headers,
            sync_client=sync_client,
        )

        # Transform response
        return self.config.transform_image_generation_response(
            model=model,
            raw_response=final_response,
            model_response=model_response,
            logging_obj=logging_obj,
            request_data=data,

View on GitHub (pinned to 6c2dcb801b)

Solutions

1. Read the embedded exception text — it distinguishes DNS vs TLS vs timeout vs proxy failures and dictates the fix.
2. Verify connectivity from the same host: curl -sS https://api.bfl.ai/ should reach the service (any HTTP response proves the path).
3. Fix proxy config (unset or correct HTTPS_PROXY) or open egress to api.bfl.ai.
4. Retry with backoff — connect errors and timeouts are frequently transient.
5. Increase the request timeout if the embedded error is a ReadTimeout.

null

null

from tenacity import retry, stop_after_attempt, wait_exponential

@retry(stop=stop_after_attempt(3), wait=wait_exponential(min=1, max=20),
       retry_error_callback=lambda e: e)
def generate():
    return litellm.image_generation(model=M, prompt=p)

When it happens

Trigger: litellm.image_generation(model="black_forest_labs/...", ...) where the initial POST to the BFL submission URL raises: ConnectError, ConnectTimeout, ReadTimeout, DNS resolution failure, SSLError, or proxy connection refusal.

Common situations: Egress blocked by firewall/security groups; wrong HTTPS_PROXY/HTTP_PROXY env vars; BFL API outage; transient network blips; self-signed TLS interception appliances; overly aggressive client timeout passed to the call.

Related errors

• Polling timed out after {max_wait} seconds
• Error from qdrant checking if /collections exist {collection
• Mavvrik FOCUS destination: register failed ({resp.status_cod
• Mavvrik FOCUS destination: failed to get signed URL ({resp.s
• Failed to fetch prompt '{prompt_id}' from API: {e}

AI-assisted analysis of BerriAI/litellm@6c2dcb801b (2026-08-15). Data as JSON: /api/errors/591d80228c3af319. Report an issue: GitHub.