How many reference images? 10 on Image 1.0, 16 on GPT Image 2.5
Image 1.0 image_urls takes 1 to 10 public HTTPS URLs; ChatGPT Image 2.5 takes up to 16 references. Text-only models reject references.

Image 1.0 accepts 1 to 10 reference or edit images in image_urls, and ChatGPT Image 2.5 accepts up to 16 image references on POST /v1/images. A model whose input_references descriptor is min 0, max 0 is text-to-image only and rejects references.
What are the limits by route?
Both routes require public HTTPS URLs.
| Route or model | Field | Limit |
|---|---|---|
| Image 1.0 | image_urls | 1 to 10 |
| ChatGPT Image 2.5 | input_references | Up to 16 |
| Text-only model | input_references | min 0, max 0, rejects references |
What counts as a valid URL?
Localhost, private-network and non-HTTPS URLs are rejected before submission. If Sume cannot fetch or mirror an input, the job fails with an error such as image_not_fetchable; check the URL is public and returns an image.
How do I send several references?
List them in image_urls on Image 1.0:
curl -X POST https://api.sume.com/v1/image-1.0/generate \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: refs-001" \
-d '{
"prompt": "Place the mug from image 1 on the table from image 2",
"image_urls": [
"https://media.sume.com/artifacts/artf_demo/mug.png",
"https://media.sume.com/artifacts/artf_demo/table.png"
]
}'Which route for many references?
Use POST /v1/images with ChatGPT Image 2.5 when you need more than 10. Read the model's input_references descriptor first, since limits vary by model.
How do I check this myself?
Confirm each URL opens in a private browser window before submitting. 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 Models
- Image quality default on Sume: low on Image 1.0, high on GPT 2.5
Image 1.0 defaults quality to low; ChatGPT Image 2.5 on POST /v1/images defaults to high when omitted. Values, and when to raise quality.
- gpt-image-2 or openai/gpt-image-2? Model ids on Sume's Image API
Sume accepts bare Image Router ids such as gpt-image-2 and nano-banana-2 as aliases for openai/gpt-image-2 style ids. The catalog lists org/slug ids.
- Kling 2.1 API sunset: 410 Gone, and the kling-3 id on Sume
Krea retired Kling 1.0 through 2.1: endpoints return 410 Gone after Sep 15, 2026. Sume lists kling-3 in its video catalog; here is what to change.
- Kling 4.0 30 seconds: Sume lists kling-3 at 4 to 15
Kling 4.0 announces 3 to 30 second generation. Sume lists kling-3 at 4 to 15 seconds; seedance-2.5 and wan-3.0 are the ids that reach 30 seconds.
Written by Sume