Canva has unveiled Brush, a vision-language model built by its research team for editing designs. The model looks at a design, understands its structure and applies the changes the user asks for, from a single line of copy to every slide in a deck. According to the company, it is up to 69% faster and 46% cheaper per task than its previous design agent. Brush now powers edits in Canva AI for Canva Pro subscribers, with wider availability to follow.
What it means: Brush arrives a week after Canva Craft, the company's first image generation model, and completes a clear strategy: Canva wants its own specialised models for its core tasks instead of relying only on general-purpose third-party models. The main reason is economic, as the company itself admits: with hundreds of millions of users, even a few cents per task become a huge cost. A smaller model trained specifically on design editing can be cheaper and faster without losing quality. For small businesses and freelancers in Greece who make their own posts, leaflets and presentations, being able to change whole designs with one instruction, for example turning a slide into a social media post, saves time. Results stay fully editable, so the user keeps control. For design studios, pressure grows on simple, repetitive jobs, while value shifts towards creative direction and brand identity. Canva says it is preparing Brand Kit support, which will let businesses keep colours and fonts consistent in automated edits.
Related service: Custom software
Related stories
- OpenAI introduces GPT-6 Cyber for aligned cybersecurity research
- NVIDIA Cosmos offers open tools and checkpoints for interactive world simulation
- Mistral Studio expands with managed agents, workflows and enterprise apps
- Google launches Gemini Omni 1.2 Flash with 15-second video and 30% lower price
- Dropbox plans to buy Tigris and launches DBX Agent Cloud for AI storage