Canva released Canva Craft, its first image model built from scratch

Canva unveiled Canva Craft on October 22, the first foundation model built and trained from scratch by Canva Research, the company's in-house AI lab. The image generation model is available to all Canva users worldwide through Dream Lab, and to developers through Canva's public API.

Craft works differently from most image models, which go straight from a prompt to pixels. It uses two stages. First, a planning model reads the prompt and writes a detailed plan of the composition: which elements appear, where they sit relative to each other, their scale, the visual hierarchy and any text that needs to show up. Then a separate rendering model turns that plan into an image at up to 2K resolution in more than 30 aspect ratios. Canva says separating planning from rendering gives more precision, consistency and control, which matters for designs where layout and typography count as much as looks.

Canva says Craft makes it the second most advanced AI lab for image generation, based on rankings from DesignArena, which it describes as the largest crowdsourced assessor of AI design. That is a company claim we could not check independently. Coverage also notes that a planned connection to third party AI assistants through Canva's MCP support has not fully launched yet.

The launch matters strategically for Canva. As its first foundation model trained in house, Craft puts the company in direct competition with image models from OpenAI, Microsoft and Google.

Why it matters for developers: a design focused image model with a public API is a new option for apps that need on-brand graphics, marketing assets or layouts with readable text, an area where general image models still struggle. Teams should test typography, editability and consistency across repeated variations before trusting launch benchmarks.

What's next: watch for API pricing and limits, the MCP integration going live, and independent comparisons against other leading image models on text rendering and layout accuracy.

Building with AI? See this week's hot AI repos and how to use each one.