Designers needing accurate text in images or exact brand colours; cheap per image on Alibaba's API, with UK access unconfirmed.

Kenning's take

What it does

Wan2.7-Image is a single Alibaba model for generating images from text and editing them with instructions or by selecting areas to change. It renders text prompts of up to 3,000 tokens in 12 languages, including formulas and tables, accepts up to nine reference images and makes up to 12 images at once. A colour palette option takes colour codes to match brand colours, and a Pro version adds 4K output. It is available through Alibaba Cloud Model Studio.

  • Text-to-image and instruction-based editing
  • Click-to-edit selected areas
  • Text rendering in 12 languages, up to 3,000 tokens
  • Colour palette control with colour codes
  • Up to nine reference images and 12 outputs at once
  • Pro version with 4K output

Who it suits. Designers, marketers and developers who need images with accurate text or exact brand colours.

Key facts

Category
AI models
Made by
Alibaba
Based in
China
Released
3 April 2026
How to get it
API on Alibaba Cloud Model Studio
Works on
API
Free plan
Model Studio free quota of 50 images, valid 90 days
API price
wan2.7-image $0.03 per image; wan2.7-image-pro $0.075 per image (international)

Reviews and write-ups

FAQ

When was Wan2.7-Image released?

Alibaba announced Wan2.7-Image in early April 2026; Arabian Reseller reported the launch on 3 April 2026.

How many languages can Wan2.7-Image write text in?

12 languages, with text inputs of up to 3,000 tokens.

How much does Wan2.7-Image cost?

$0.03 per image on Alibaba Cloud Model Studio, or $0.075 for the Pro version.

Facts from the maker's own pages and the sources listed. Checked 2026-10-11.