Canva 发布首个自研图像模型 Craft

Canva today released Craft, its first-ever AI model for images that was built specifically for design, rather than photography.

“Craft is the first foundation model built and trained from scratch by Canva Research, bringing together years of AI research and investment to deliver a deeper understanding of design, results that more closely match what you imagined, and the freedom to easily edit and refine everything you create,” the Canva announcement explains. “It debuted as the fourth most advanced image generation model on DesignArena [behind three OpenAI models, GPT-Image-2.5 Sunburst, GPT-Image-2.5 Flare, and GPT Image 2, delivering industry-leading visual quality and accurate text rendering to support a wide range of creative tasks.”

Craft creates 2K images at any aspect ratio with precise control, editable outputs, and high quality. It was trained for design workflows using millions of human-made templates from Canva’s content library, which it says spans professionally designed posters, social posts, presentations, infographics, menus and typography-rich layouts. It can create photorealistic images, but it was made for social content, marketing assets, lifestyle and editorial photography, illustration and digital art, people and characters, nature and environmental imagery, and creative and conceptual work.

Craft is available now in the Canva app–so you will need a paid subscription–and to developers through its public API and, soon, the Canva MCP Server so agents can use it to automate tasks. And Canva says it will soon release two variants, Craft-fast and Craft-ultrafast, that are designed to deliver high-quality image generation at a fraction of the cost of leading frontier models.