Canva Launches First AI Image Model, Canva Craft

Canva released its first AI image generation model, Canva Craft, for all users.

Canva has launched its first in-house AI image generation model, named Canva Craft. This new model is now available to all users on the company's platform, accessible via the Dream Lab tab, and will also be available through the Canva API. Canva Craft is designed to compete with existing models from companies like OpenAI, Microsoft, and Google, promising improved, more customizable performance at potentially lower costs for Canva.

Canva Craft utilizes a two-stage architecture. In the initial stage, it interprets user prompts to create a detailed plan, including spatial relationships, relative scale, and visual hierarchy. The second stage then generates an image with up to 2K resolution in various aspect ratios. This approach aims to produce images with accurate object details and legible text, allowing for precise edits without affecting other parts of the image. The model was trained using human-designed templates from Canva’s library, rather than photographs.

This release signifies Canva's deeper dive into AI, with the company noting that the new model will help save API costs while offering better quality and speed control. Researchers from Leonardo.ai, a media generation platform acquired by Canva, contributed to the development of Canva Craft. The company also plans to release two cost-efficient versions, Craftfast and Craft-ultrafast, in the near future.

Find the right AI tool for your business

Chat with Insta and get matched to the right tool in seconds.