Prompt-only or edit-first
Use it for a brand-new image or pass one or more public reference image URLs when you want to restage, restyle, or refine an existing visual.
Higher-cost 2K generation and edit tier for teams that want the WAN 2.7 Pro route directly inside UlazAI.
This route gives you one 2K image model that works for both fresh generations and edit-first workflows, so you can test prompts quickly and then reuse reference images without switching tools.
Use it for a brand-new image or pass one or more public reference image URLs when you want to restage, restyle, or refine an existing visual.
The current provider docs expose a 2K request path, so the Image Studio setup stays intentionally simple instead of pretending the model has modes we have not verified.
Provider pricing is 12 credits per image, so UlazAI billing stays at 12 credits with no rounding delta.
Send a prompt to the same async image route. For edits, add one or more image URLs and keep the quality at 2K.
Current docs show prompt, optional input_urls, fixed 2K resolution, and optional region metadata. Image Studio uses the minimal safe payload rather than overfilling speculative inputs.
Like the other UlazAI async image models, you can poll the task record endpoint or wait for your callback and then read the final result URL.
| Model | UlazAI credits | Provider rate | Best fit |
|---|---|---|---|
| WAN 2.7 Image Pro | 12 | 12 | Premium product shots, refined ad creative edits, sharper commercial deliverables, and higher-value prompt iterations. |
| WAN 2.7 Image | 5 | 4.8 | Higher-value commercial image work with the same 2K async flow. |
Yes. The current documented route covers both prompt-only generation and edits when you pass reference image URLs.
Provider pricing is 12 credits per image, so UlazAI billing stays at 12 credits with no rounding delta.
Yes. The same model is available through Image Studio for testing and the async image API for production integration.
We currently expose this route with provider-managed framing because the public provider docs show 2K requests clearly but do not document a verified ratio matrix we want to promise publicly.
Open the model in Image Studio, test a prompt-only render or upload a reference image for editing, and keep the workflow on the same async stack when you move to API.