Canva launches Craft, a two-stage AI image generation model
Canva unveiled Canva Craft, its first in-house AI image model, which plans a design before rendering it to improve accuracy on text and object placement. The model supports resolutions up to 2K and over 30 aspect ratios, and was trained on millions of templates from Canva's content library. It currently ranks fourth on the Design Arena leaderboard, behind three OpenAI models.
GoKawiil's interpretation of the reporting above, not reported fact.
The launch signals Canva's push to build proprietary AI infrastructure rather than relying solely on partner models, which could reduce dependence on OpenAI, Google, or Microsoft over time. Its ranking behind OpenAI's offerings suggests Canva still has ground to cover to match leading image generators on quality. The focus on moderation and safety controls may reflect growing regulatory and user scrutiny of AI-generated content.
- Canva Craft uses a two-stage plan-then-generate architecture for more precise image creation.
- It supports up to 2K resolution and more than 30 aspect ratios, trained on Canva's own template library.
- The model ranks fourth on the Design Arena leaderboard, trailing three OpenAI image models.
Source: theverge.com — Emma Roth, 2026-10-22
Published there as: “Canva’s first AI image model plans designs before generating them”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.