Canva Craft works in two stages. In the first stage, it interprets user prompts and creates a plan detailing spatial relationships, relative scale, visual hierarchy, and text generation information. In the second stage, it generates images up to 2K resolution in up to 30 different aspect ratios for various platforms. Canva said this approach allows for accurate object details and legible text, and users can make precise edits without altering other image parts.
The company trained the model on human-designed templates from Canva's library, including posters, social posts, presentations, infographics, menus, and typography-rich layouts. This design-focused training means generated images can be converted into layered designs using Canva Magic, the company's image-to-design transformer. Last year, Canva released a model that understands designs and layers.
Canva has previously used image models from other providers for generation, so the new model could help reduce API costs. The company stated it also allows for better quality and speed control. Researchers from Leonardo.ai, which Canva acquired in 2024, contributed to the new model.
Canva plans to support image generation in third-party chatbots like ChatGPT and Gemini through Canva's MCP. Additionally, the company is developing two cost-efficient versions, Craftfast and Craft-ultrafast, to be released soon.
In business context, Canva was on track for $4 billion in revenue last year, though it cut its growth forecast for 2026, reportedly causing its valuation to fall from $42 billion to $35 billion according to external analysts.
