Canva launches its first image‑generation model
google meta openai
| Source: TechCrunch | Original article
Canva has launched its first image‑generation model, Canva Craft, now available to all users and touted for more customizable performance at lower cost.
Canva has rolled out its first proprietary image‑generation model, Canva Craft, making the tool available to every user of the design platform. Developed by Canva Research, the in‑house AI lab, Craft is positioned as a design‑centric alternative to image generators from Google, OpenAI, ByteDance and Meta. The model plans a composition before rendering, a workflow that Canva says improves text placement, layout consistency and granular control over visual elements. It can produce images up to 2K resolution across more than 30 aspect ratios, and the company promises “better, more customizable performance at lower costs” for its broad user base.
The launch marks a shift for Canva from relying on external providers to building core creative AI capabilities internally. By tailoring the model to design tasks, Canva aims to reduce the friction that designers often encounter when adapting generic generators to precise layout needs. The move also pushes Canva up the rankings of AI labs for image generation, according to DesignArena, the largest crowdsourced assessor of AI design, making it the world’s second‑most advanced lab in this niche.
Industry observers will watch how quickly creators adopt Craft and whether the lower‑cost promise translates into measurable savings for Canva’s largely subscription‑based clientele. Competition will likely respond, with rivals potentially accelerating their own design‑focused model development. Further updates from Canva on integration depth—such as real‑time editing, API access or enterprise‑grade licensing—could signal how the company plans to leverage Craft beyond its consumer platform and into broader creative workflows.
Sources
Back to AIPULSEN