Particle.news

Canva Debuts Magic Layers to Turn Flat Images Into Editable Designs

Powered by Canva’s proprietary design model, the tool reconstructs elements into separate layers for faster, precise edits.

Overview

  • Magic Layers launches in public beta today in the US, UK, Canada, and Australia, with broader availability planned but no timeline announced.
  • It ingests single-page PNG or JPEG files, including AI-generated images, and separates layouts into movable objects and live text boxes.
  • Extracted components behave like native Canva elements, allowing color changes, resizing, removals, upscaling, Magic Edit, and real-time collaboration.
  • The feature runs on the Canva Design Model introduced in October 2025 alongside a broader stack that includes models from OpenAI and Anthropic.
  • Use requires a Canva subscription, and journalists characterize the tool as a practical fix for current generative-AI editing limitations.