Canva AI tools have been lacking or, at least, trailing the big name tools like ChatGPT, Gemini, Claude, etc.
This may help them get back in the game.
They have developed and trained their own image model called Canva Craft, and it’s good.
The current iteration can handle at least 30 different aspect ratios, outputs 2K images, and rivals the best image models out there (GPT Image 2.5, Nano Banana 2.1, etc.)
Their model uses a 2-stage architecture that starts with a planning phase and only generates the image after it has reasoned about its structure. This approach seems to deliver greater precision, consistency and control.
One of the biggest issues that I’ve had with image generation in Canva is text on images. From the output that I’ve gotten, they’ve fixed those issues. They say that text can be created in 55 languages and prompts can be written in any language.
Pair this model with existing Canva tools and Canva will be hard to beat for creative work.
Canva Craft is available to all Canva users globally today.
(go try it)
Watch this video A Demo:
.png?width=600&height=496&name=Next-Level%20AI%20Workshops%20(2000%20x%201000%20px).png)
You’ve Read Enough About AI. Time to Do Something With It.
Hands-on workshops for people who want results, not theory.
Comments