Wan 2.7 Image Pro
What is Wan 2.7 Image Pro?
Wan 2.7 Image Pro is the highest-quality image model in Alibaba's Wan line. It exposes the same parameter shape as Wan 2.7 Image (up to nine reference images, thinking mode for text-only runs, seven aspect ratios, 5000-character prompts) and pushes output up to 4K. The model auto-selects between text-to-image and editing based on whether image_urls is set. Use it for print, packaging, OOH, and any final asset where quality matters more than turnaround.
Key features of Wan 2.7 Image Pro
Four features cover what changes versus the standard Wan 2.7 Image.
Up to 4K output
The headline upgrade over the standard model. Same prompt, same references, same seed, higher resolution. Useful for print, packaging, and large-format displays.
Up to 9 reference images
Same nine-reference cap as the base model. Useful for brand toolkits, character casts, and multi-style alignment in one call.
Thinking mode on text-only runs
`thinking_mode: true` adds a reasoning pass before drawing. On by default. Useful for dense, multi-subject prompts where composition matters.
Editing preserves input aspect ratio
When references are present, output keeps the source aspect ratio. Useful for product retouching and character restyling where the layout must not shift.
Best for
Print and large-format
4K marketing creative, packaging, OOH, and any frame that needs to survive heavy crop and zoom.
Complex briefs
Thinking mode on text-only generations for harder scenes with many elements or subjects.
Brand toolkits
Up to nine references in a single request keeps colour, character, and layout consistent across a series.
Premium stills
Final-quality output when longer renders are acceptable and quality matters more than throughput.
Multi-reference editing
Up to nine reference images for style and subject alignment on existing assets.
High-detail composition
Highest quality in the Wan image line. Useful when 2K plateaus.
API examples
Call Wan 2.7 Image Pro from any language by POSTing to /v1/tasks. Full parameter docs live at docs.unifically.com/models/image/alibaba/wan-2.7-pro-image.
curl -X POST https://api.unifically.com/v1/tasks \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{
"model": "alibaba/wan-2.7-pro-image",
"input": {
"prompt": "A flower shop with exquisite windows and flowers on display",
"aspect_ratio": "16:9"
}
}'
Successful submission returns a task_id. Poll GET /v1/tasks/<task_id> or set a callback_url on the request to receive the finished result.
FAQs
People also ask
Wan 2.7 Image Pro is the highest-quality image model in the Wan line. It supports text-to-image, editing, and image-guided generation with up to nine reference images, seven aspect ratios, thinking mode for text-only runs, and 4K output. Same parameter shape as the standard model, tuned for quality over speed.
Thinking mode (thinking_mode: true) adds a reasoning pass before drawing on text-only generations. It is on by default. Not used when reference images are supplied.
Up to nine. Useful for steering style, subject, composition, or character consistency across a series.
Pick Pro when you need 4K output and the highest quality in the line. Pick the standard model when 2K is enough and shorter turnaround matters.
Yes. Run drafts on the standard model, lock the keepers with a seed, then re-render the finals on Image Pro.
Yes. Pass an integer seed alongside the prompt and any references. Re-sending the same inputs and seed returns the same image. Useful for locking a draft on the standard model, then re-rendering at 4K on Pro.
5000 characters for the main prompt. The aspect ratio in editing mode is preserved from the input image.
Guides and comparisons
Review pricing, limits, and tested outputs before running this model.
