Canva, the Australian graphic design platform, has launched Magic Layers, a feature that converts flat, static images into fully editable, multi-layered designs within its editor.
The tool addresses a recurring limitation of AI-generated images, which typically export as static JPEG or PNG files that lose editable text, distinct objects and layout structure in the process.
Magic Layers analyses an entire design and reconstructs it as a set of structured layers: text is restored as live, editable text boxes, objects are separated into individual elements, and backgrounds are preserved behind foreground content so the original layout remains intact.
"We believe AI should spark creation, not bring it to a halt," Canva said in its announcement.
The feature works on images imported from external AI generators as well as on existing projects already inside the Canva editor.
Magic Layers is built on the Canva Design Model, the company's proprietary AI system, which Canva says has produced hundreds of millions of editable presentations, documents and social posts through its suite of AI tools.
Related reading
- Canva embeds designs in ClassDojo to cut tool-switching for teachers
- Canva acquires MangoAI and Cavalry as the design platform expands AI and enterprise-grade capabilities
- Canva's latest acquistion expands its professional design stack
The Design Model also powers Canva's integrations with ChatGPT, Claude and Microsoft Copilot.
Magic Layers is available to users in the Canva editor now.
The recap
- Canva launches Magic Layers to convert flat images into layers
- Design Model has generated hundreds of millions of editable assets
- Users can import images and try Magic Layers in editor today