Key points
- Canva Craft uses a two-stage process to plan spatial relationships before generating images up to 2K resolution.
- The model is trained on human-designed templates rather than raw photographs, enabling better text legibility.
- Access is available to all users via the Dream Lab tab and through the Canva API for integration.
The short version
- Canva launched Canva Craft, its first in-house image generation model, available to all platform users.
- The system employs a two-stage architecture that creates a design plan before rendering the final image.
- Training data consists of human-made templates from Canva’s library, distinguishing it from photo-based models.
The details
Canva has released Canva Craft, its first proprietary image generation model, making it available to all users on the platform. According to TechCrunch, the design tool’s research team developed the model to compete with offerings from Google, OpenAI, ByteDance, and Meta. The system operates in two distinct stages. First, it interprets user prompts to create a structural plan that defines spatial relationships, relative scale, visual hierarchy, and text generation details. Second, it generates the final image at up to 2K resolution across more than 30 aspect ratios.
This approach allows for precise edits without altering other parts of the image, according to Canva. The model generates images with accurate object details and legible text. Canva noted that it trained the model using human-designed templates from its library, including posters, social posts, presentations, infographics, menus, and typography-rich layouts, rather than using photographs. This training method enables generated images to be converted into layered designs using Canva Magic, the company’s image-to-design transformer.
Canva Craft is accessible from the Dream Lab tab within the platform and through the Canva API. The company also plans to support image generation in third-party chatbots like ChatGPT and Gemini via the Model Context Protocol. Researchers from Leonardo.ai, which Canva acquired in 2024, contributed to the new model. Canva stated that developing its own model helps save API costs and provides better control over quality and speed.
The context
The release follows Canva’s broader push into AI features, including prompt-based editing capabilities launched in April. The company previously released a model that understands designs and layers, allowing users to modify parts of an image while leaving the rest untouched. Canva Craft is currently ranked fourth on Design Arena’s leaderboard for AI image generation models, sitting behind several versions of OpenAI’s GPT-Image models.
Canva is developing two cost-efficient versions of the model, called Craftfast and Craft-ultrafast, which will be released soon. The company emphasized that the model includes robust input and output moderation, privacy and safety controls, and reporting tools. This launch occurs as Canva adjusts its financial outlook, having cut its growth forecast for 2026, which reportedly caused its valuation to drop from $42 billion to $35 billion, according to external analysts.
What to watch
- The upcoming release of Craftfast and Craft-ultrafast, which are described as cost-efficient versions of the current model.
- Integration of Canva Craft into third-party chatbots like ChatGPT and Gemini through the Model Context Protocol.
- Performance comparisons on Design Arena’s leaderboard as new models from competitors like OpenAI and Google emerge.
What it means for organisations
Organisations using Canva for marketing materials should assess whether the new model’s output meets their brand guidelines. Teams usually need to verify that generated text is legible and that spatial relationships align with corporate identity standards. Because the model is trained on design templates, it may produce different results than photo-based AI tools. Security teams should review the API access policies if developers plan to integrate Canva Craft into internal workflows. Management should be informed about the shift from third-party image providers to Canva’s proprietary model, which may affect long-term vendor dependencies and cost structures.
What to do and how to stay safe: Canva
- Review current image generation workflows to determine if Canva Craft’s template-based training aligns with your content needs.
- Test the model’s output for text accuracy and spatial consistency before using it for official marketing materials.
- Check your organisation’s API usage policies if you plan to integrate Canva Craft into automated design pipelines.
- Monitor the release of Craftfast and Craft-ultrafast for potential cost savings on high-volume image generation tasks.
Step-by-step guide: Secure Home Routers for Small Business Networks
General security guidance from the TechSec Headlines newsroom. It is not confirmed advice from the organisations named in this story.
Frequently asked questions
How does Canva Craft differ from other AI image models?
Canva Craft uses a two-stage architecture that plans design elements before generating images, and it is trained on human-made templates rather than photographs.
Is Canva Craft available to free users?
Yes, Canva Craft is available to all users on the platform, accessible through the Dream Lab tab.
What resolution and aspect ratios does Canva Craft support?
The model generates images at up to 2K resolution and supports more than 30 different aspect ratios for various platforms.
