
What happened
Canva launched Canva Craft, an AI image model that plans a design before generating it, producing images up to 2K resolution in over 30 aspect ratios.
Why it matters
The two-stage system aims for greater precision and consistency, helping create complex images with legible text and precise object placement, per Canva.
WHO IT HITSDesigners and marketers who create social media posts, presentations, and infographics can now use Canva Craft to generate images with more control, potentially reducing manual editing.
Summaries like this, every morning.
Canva's launch of Canva Craft builds on its earlier AI push. In April, the company introduced prompt-based editing capabilities and brought layers to AI-generated images, letting users modify part of an image while leaving the rest untouched. Canva says it trained Craft on "millions of human-made templates" from its content library, which includes social media posts, presentations, infographics, and menus. Currently, the model is ranked fourth on Design Arena's leaderboard, a crowdsourced benchmark for AI designs, sitting behind OpenAI's GPT-Image-2.5 Sunburst, GPT-Image-2.5 Flare, and GPT-Image-2. The company also plans to allow AI agents to plug into Canva Craft through the open Model Context Protocol, and will roll out more efficient Craft-fast and Craft-ultrafast models in the future.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
Canva released Canva Craft, its first image generation model, available to all users via the Dream Lab tab and…

Rerunning the same images and prompts that broke Gemini 3.5 Flash Lite, the developer found Gemini 4 Argon enc…

KDDI, Okinawa Cellular and KDDI Digital Life added ChatGPT 5.5, ChatGPT 5.5 Pro and ChatGPT 5.5 Thinking to po…

IronSight AI and Fast Iron announced a partnership integrating IronSight's image recognition into Fast Iron's…

Fontify's post lays out four observable contracts — input, style, layout, output — each paired with a failure…

He, Zhang, Ren and Sun at Microsoft Research showed that making each block learn only a residual $F(x) = H(x)…
