Canva launches Craft, its first in-house image model, which plans a layout before it generates
Trained on human-made design templates, Craft aims for readable text and editable posters, slides and social graphics.
Design platform Canva has released Canva Craft, its first image generation model developed in-house. It competes with models from Google, OpenAI, ByteDance and Meta. Craft works in two stages. It first interprets a prompt and plans the image: where objects go, their relative scale, the visual hierarchy, and any text. It then generates the image at up to 2K resolution in 30 aspect ratios. Canva says this approach produces more accurate details and readable text, and lets users edit specific elements without changing the rest of the image. Instead of photographs, the model was trained on human-made templates from Canva's library, including posters, social posts, presentations, infographics, menus and typography-heavy layouts. As a result, Craft's output can be turned into layered, editable designs with Canva Magic, the company's image-to-design tool. Craft is available to all Canva users through the Dream Lab tab and will also be offered through the Canva API. The company also plans to support image generation in third-party chatbots. In a post on X, Canva said Craft makes it the second most advanced AI lab for image generation on the DesignArena leaderboard. The launch shows design software makers building their own models instead of relying only on general-purpose image generators.