Canva Launches Craft, Its First In-House Image Generation Model

Canva has released Canva Craft, its first in-house image generation model. The tool was developed by the company's research team to compete with models from Google, OpenAI, ByteDance, and Meta. Researchers from Leonardo.ai, which Canva acquired in 2024, contributed to the development of the new system.

The model operates in two stages, first interpreting prompts to create a plan with spatial relationships and text generation information, then generating the final image. It supports up to 2K resolution in up to 30 different aspect ratios and can render text in 55 languages. Canva trained the model using human-designed templates from its library rather than photographs. Users can make precise edits without changing other parts of the image, and generated images can be converted into layered designs using Canva Magic.

On the DesignArena platform, Craft placed fourth overall and first in portraits and landscapes. The model is currently available to all users via the Dream Lab tab and the Canva API. Canva plans to support image generation in third-party chatbots like ChatGPT and Gemini through its MCP later. Two cost-efficient versions, Craft-fast and Craft-ultrafast, are expected soon. Canva’s valuation reportedly dropped from $42 billion to $35 billion after cutting its growth forecast for 2026.