Can the Sume Image API output SVG? Not today; formats you get
No: output_format svg is in the schema for parity, but no catalog model advertises it, so the request is rejected. Sume returns png, jpeg or webp.
No. output_format: "svg" appears in the public vocabulary, but the repo's own deferral table says no v1 model advertises it, so the catalog check rejects it per model. The formats you can request on Sume today are png, jpeg and webp where the row lists them. recraft/recraft-v4 is webp-only, qwen/qwen-image-layered is png, and ideogram/ideogram-v4.5 lets the provider choose, so it does not accept the field.
Formats by row
From the repo's catalog capabilities.
| Sume model id | output_format values |
|---|---|
| openai/gpt-image-2.5 | png, jpeg, webp |
| google/nano-banana-2.1 | png, jpeg, webp |
| recraft/recraft-v4 | webp |
| qwen/qwen-image-layered | png |
| ideogram/ideogram-v4.5 | Provider-selected; field not accepted |
| higgsfield/soul | Provider-selected (PNG) |
Workarounds
For logos and icons, generate a clean raster at the largest tier and vectorize it in your own tool. Ask for flat colors, a plain background and no gradients to make that step easier. For transparent stills, openai/gpt-image-2.5 accepts background: transparent.
A request with svg on a row that does not list it returns 400 unsupported_parameter. Test one call before a batch.
curl -s https://api.sume.com/v1/images \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"openai/gpt-image-2.5","prompt":"flat vector style leaf icon, solid colors","background":"transparent","output_format":"png"}'Watch for
- Vectorizing raster art can lose fine detail.
- Brand marks need a human check for similarity and rights.
Sources
Related posts
More in Developers
- GET /v1/balance returns available_credits: a TypeScript gate
Sume's balance read returns whole credits in available_credits, a funded or empty state, and expiring_soon_credits. Gate a job on it in TypeScript.
- Go 1.27.2 synctest: test a 30-second Sume poll loop in milliseconds
Test a Go poll loop for Sume video jobs with testing/synctest: 90 seconds of fake sleeps and a 20-minute deadline pass in under a second, no sockets.
- Image edit reference limits on Sume: 16 on GPT Image 2.5, 10 elsewhere
ChatGPT Image 2.5 takes up to 16 input_references; most edit rows take 10, Ideogram 4.5 takes 5. Imagen, Recraft and Qwen Image Max take none.
- GPT Image 2.5 custom size: 1920x1080 fails, use 1920x1088
Edges must be multiples of 16, so 1080 is out. 1920x1088, 1280x720 and 3840x2160 pass the Sume rules; 1080x1350 does not, 1088x1360 does.
Written by Sume