Canva Craft
Canva’s image generation model creates design visuals from text prompts in Dream Lab and through a preview API.
MAI Image 2.7
microsoft.ai
Microsoft AI’s preview image editing model revises scenes, lighting, lettering, and existing visual assets.
Visit website
Pricing
Not confirmed
Available on
Web via MAI Playground
Category
Ai Image Tools
What is MAI Image 2.7?
MAI Image 2.7 edits existing images for creators and design teams. Microsoft offers it in preview and directs users to MAI Playground to try editing.
Key features
Scene restyling
Change the visual style of an existing scene.
Relighting
Adjust scene lighting and mood.
Text and symbol edits
Revise lettering and symbols while keeping them integrated into the image.
Restoration and enhancement
Restore or enhance existing visual assets.
Product and object edits
Place a product in a new setting or edit an object with attention to surrounding lighting and environment.
Who is it for?
Product designers
Explore a different setting for an existing product visual.
Creative teams
Compare lighting directions for a campaign asset.
Content editors
Request a lettering revision within an illustration.
How to use MAI Image 2.7
- Visit the official announcement and follow its MAI Playground link.
- Look for the MAI Image 2.7 editing entry.
- Prepare an existing asset and a short description of the desired edit, then follow the editing entry’s instructions.
- Review the result against your original brief.
MAI Image 2.7 pricing
Not confirmed
Pricing for the MAI Image 2.7 preview is not confirmed. Consult Microsoft’s official model pricing information for updates.
Check current pricing ↗What to know before you use it
- The model is available in preview, and MAI Playground is a limited preview environment.
- Text-to-image generation is described as coming soon.
Frequently asked questions
Can it edit individual objects?
Yes. Microsoft describes object edits that account for surrounding lighting and environment.
Where else can I try editing?
Microsoft also directs users to Arena and Artificial Analysis.