Qwen Image 2.1 Pro: what exists, and which Qwen ids Sume lists
People search for Qwen Image 2.1 Pro. Its model card lists no Pro variant; Sume's Image API lists qwen/qwen-image and qwen/qwen-image-max.

There is a Qwen-Image 2.1 model card on Hugging Face, and the page I read documents no "Pro" variant. Sume's Image API does not list Qwen-Image 2.1 either. It lists two Qwen models, qwen/qwen-image and qwen/qwen-image-max, so if you searched for "Qwen Image 2.1 Pro" and want a hosted Qwen on Sume, those two are the ids to try.
What the model card documents
The Qwen-Image-2.1 card describes a model with 7B parameters in its visual generation component, text-to-image and editing in one model, and up to 10 reference images. It lists a maximum output of 2752x1536 pixels at 16:9 and seven standard ratios from 1:1 to 9:16. Its license is the Qwen Research License Agreement. Third-party pages mention a Pro variant, but I could not confirm it on the vendor page, so I do not describe it here.
| Question | Answer |
|---|---|
| Is there a documented Pro variant on the model card? | No Pro variant mentioned |
| Maximum output on the card | 2752x1536 (16:9) |
| Reference images on the card | up to 10 |
| Qwen ids in the Sume image catalog | qwen/qwen-image, qwen/qwen-image-max |
| Custom pixel image_size on Sume Qwen | yes, per the docs, as a tier or width and height object |
Calling a Qwen model on Sume
Pick one of the two ids and read its catalog row for the exact ratios, quality values and reference range, since each model accepts only what its descriptors list. Anything outside them returns 400 unsupported_parameter instead of being dropped.
curl "https://api.sume.com/v1/images/models/qwen/qwen-image-max/endpoints" \
-H "Authorization: Bearer $SUME_API_KEY"
curl -X POST "https://api.sume.com/v1/images" \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"qwen/qwen-image-max","prompt":"product shot of a bamboo toothbrush on a stone ledge, soft morning light","aspect_ratio":"16:9"}'If you need the 2.1 weights
The Qwen-Image-2.1 checkpoint is downloadable under its research license. If you want exactly that checkpoint, run it on your own hardware and check the license for commercial use first. Sume does not serve it, and the docs do not promise that a given Qwen id maps to a specific upstream release.
Choosing among the options
Use qwen/qwen-image-max when you want the larger of the two Qwen entries and are happy to read the catalog for its limits. Use sume/auto if you do not care which family runs; Sume never discloses which one served an auto request. Pin a model id when you need repeatable looks across a batch.
How to confirm what is live
Do not trust any post, including this one, over the live catalog:
GET /v1/images/modelslists the ids that exist today.- The endpoint record shows the price and the descriptors.
- A retired id keeps working and runs as its successor, so check the job's stored model.
Sources
Related posts
More in Models
- Remove an object from a photo without a mask: Ideogram 4.5 prompt edit
Remove a bin, cable or stray person from a photo with no mask: send it to ideogram/ideogram-v4.5 and name the object. $0.0375 to $0.275 per image on Sume.
- Replace one prop in a video with AI: bottle to apple edit prompt
Swap one object in a finished clip with gemini-omni-flash-1.1 on Sume: the docs example prompt, fields you cannot set, and an 8-second price.
- Seedance pixel sizes by ratio on Sume: 864x496 at 480p, 1920x1080
Sume prices Seedance by video tokens counted from fixed pixel sizes per ratio and resolution. The size table, and why 480p 16:9 is 864x496 and not 854x480.
- Silent clips after Sora: which Sume models take generate_audio false
Gemini Omni Flash 1.1 rejects generate_audio false; Kling 3 prices audio on at $0.21 a second against $0.14 off; recast and motion transfer keep source sound.
Written by Sume