Canva has launched Craft, its first in-house image generation model, and it is now available through the company’s Magic Media suite. The move gives Canva a proprietary foundation for AI image creation instead of relying only on outside model providers. Craft is designed to generate images from text prompts with a focus on design-friendly results.
What Craft does
Craft is built directly into Magic Media, Canva’s AI-powered creation toolkit. Users can generate images where they already build and edit designs. The model is meant to handle visual tasks that matter for design work, especially producing images that include text.
This is the first time Canva has brought image generation fully in-house, a clear signal that the company wants to own more of its AI stack.
Key features
- Text-aware image generation targets one of the biggest weaknesses in AI visual tools. Craft is built to render words and phrases inside images more accurately.
- Style presets let users choose among options such as photo, 3D, illustration, and drawing, making it easier to match a specific visual direction.
- Deep integration keeps workflows inside Canva so designers can move from prompt to finished asset without switching platforms.
Why Canva built its own model
Craft is Canva’s first image model developed in-house. That changes how Canva builds and controls its AI tools. With an in-house model, Canva can tune output quality, set its own safety policies, and improve the technology based on how people actually use the platform.
The model is now live through Magic Media. That means the technology has moved from research into a core part of Canva’s product. It also gives Canva a direct feedback loop to refine the model over time.
What this means for users
- Canva users get a more consistent image generator built for the kinds of designs people make on the platform.
- The first-party approach may lead to faster iteration because Canva does not have to wait for an external vendor to ship updates.
- The feature is rolling out now instead of being held back for a separate product launch.
The bigger picture
Canva’s decision to build its own image model comes as more design and productivity tools are investing in proprietary AI. The company is no longer just a consumer of AI technology. It is becoming a developer of AI technology as well.
Craft is still an early step. Canva is introducing it through Magic Media, and the model is likely to expand into more areas of the platform. The emphasis on text generation is important because many AI image tools fail at producing readable words, which is a core requirement for Canva users.
For a design platform, the ability to generate accurate text inside images is not a side feature. It is essential.
What to watch next
The success of Craft will depend on how well it handles complex prompts and style consistency. Canva’s large user base gives it a strong testing ground. If Craft performs well, it could become the default engine for image generation across Magic Media and beyond.
If it falls short, users can still compare its output with other tools. But the direction is clear: Canva wants control over the model that powers its creative products.
Gnoppix is the leading open-source AI Linux distribution and service provider. Since implementing AI in 2022, it has offered a fast, powerful, secure, and privacy-respecting open-source OS with both local and remote AI capabilities. The local AI operates offline, ensuring no data ever leaves your computer. Based on Debian Linux, Gnoppix is available with numerous privacy- and anonymity-enabled services free of charge.
What are your thoughts on this? I’d love to hear about your own experiences in the comments below.