Canva launches Craft, its own AI image model
Why it matters
Canva launched Craft on Oct 22, the first foundation model built and trained from scratch by Canva Research, and its first image generation model built specifically for design. Instead of generating pixels straight from a prompt, Craft uses two stages: an image planning model writes a detailed plan of the composition (which elements appear, their placement, relative scale, visual hierarchy and any text), then a dedicated rendering model turns that plan into an image at up to 2K resolution in more than 30 aspect ratios. Canva says it debuted as the fourth most advanced image generation model on DesignArena, a crowdsourced design leaderboard, making Canva Research the second most advanced lab for image generation; it ranks #1 for People & Portraits (out of 87 models) and Landscapes & Nature (out of 85), #2 for Abstract & Patterns and #4 for Products. Unlike models trained mostly on photographs, its training data includes millions of human-made Canva templates (posters, social posts, presentations, infographics, menus), and it was fine-tuned with feedback from Canva's professional designers, who judged more than 25,000 image pairs against leading models. It takes prompts in any language and can write text in images in 55 languages. Craft is available now to all Canva users in Dream Lab and through Canva's public API; Craft-fast and Craft-ultrafast, designed to deliver high-quality images at a fraction of the cost of leading frontier models, are upcoming. Caveat: the rankings come from DesignArena's live crowdsourced leaderboard, as reported by Canva, and can change.
Read the source: Canva (opens in a new tab)
All hits from 23 Oct 2026More ModelsGet the daily hits by email