FLUX image URL expires after 10 minutes: what Sume returns
Black Forest Labs says generated FLUX image URLs expire after 10 minutes. Sume returns Sume-hosted signed URLs; here is how to fetch and keep the file.

Black Forest Labs' docs say generated image URLs expire after 10 minutes, so its API results must be downloaded quickly. On Sume, FLUX.2 results come back as Sume-hosted signed URLs, and Sume's docs do not say how long those URLs stay valid, so download the file you need to keep.
Vendor facts read 2026-09-30 from Black Forest Labs; Sume facts from the Image API docs.
What does Black Forest Labs say?
Its generating-images page describes an asynchronous flow: submit a request, then "poll the endpoint using the polling_url" until the result is ready. It also states "Generated image URLs expire after 10 minutes".
What does Sume return instead?
For POST /v1/images, images return as Sume-hosted URLs (data[].url, signed) instead of inline base64. The docs say Sume already mirrors generated media, and the Image 1.0 page says raw provider URLs are not part of the public result contract. black-forest-labs/flux.2-pro is one catalog id. If a request runs past the 30-second wait it returns 202 and you read the result in the standard job shape from GET /v1/jobs/{id}/result.
| Question | Black Forest Labs API | Sume /v1/images |
|---|---|---|
| Result URL | Expires after 10 minutes | Sume-hosted, signed; lifetime not documented |
| Flow | Submit, poll polling_url | 200 with images or 202 job |
| Provider URL exposed | Not covered by this quote | Not in the Image 1.0 result contract |
How do I keep the file?
Fetch the URL and store the bytes in your own storage right after the job completes.
curl -sX POST https://api.sume.com/v1/images \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"black-forest-labs/flux.2-pro","prompt":"Studio photo of a leather bag"}' \
| jq -r '.data[0].url' | xargs curl -sSo bag.pngWhat is not documented?
Sume's docs do not give a lifetime for signed URLs, and this post makes no claim either way; copying the asset is the cautious habit. If you get 202, the command above will not work; poll the job first.
How do I check this myself?
If you build against Black Forest Labs directly, download within the 10 minutes it states. If you build against Sume, treat the returned URL as a delivery link and copy the file into your own bucket as part of the same worker that receives the result. The linked docs pages and the catalog endpoint show the current values, and this post reflects them as of 2026-09-30.
Sources
Related posts
More in Developers
- frame_images plus input_references in one request: which wins?
When a Sume video request carries both frame_images and input_references, frame_images wins and the job runs as image-to-video. What that means for references.
- Gemini video understanding API vs Sume video inspect stills
Gemini's API now has agentic video understanding. Sume's video inspect is narrower: probe facts, 8 stills and optional STT you pass to your own model.
- Gemini CLI MCP timeout default vs Sume's 55-second jobs_wait
Gemini CLI's MCP timeout defaults to 600,000 ms. Sume's jobs_wait holds at most 55 seconds per call: repeat it on wait_slice_expired, never resubmit.
- Gemini CLI keeps the OAuth refresh token: what Sume's MCP does
Gemini CLI 0.62 retains the OAuth refresh token on refresh. Sume's MCP token endpoint accepts only authorization_code, so expect a fresh sign-in, not a refresh.
Written by Sume