Design platform Canva released its first image generation model, Canva Craft, on October 22. Developed by its internal research team, the model aims to compete with similar offerings from companies like Google, OpenAI, ByteDance, and Meta. Canva stated that the new model will be available to all users, promising improved performance, higher customizability, and lower costs. The model supports image generation up to 2K resolution and offers 30 different aspect ratios. Canva claims that by training with human-designed templates from its design library, Canva Craft can generate images with accurate details and clear text, and allows for precise user editing. The model has launched on the Canva platform under the Dream Lab label and will be available via the Canva API and third-party chatbots such as ChatGPT and Gemini. Canva expects the new model to help the company save on API costs and enhance its control over platform quality and speed. According to external analyst reports, Canva's revenue reached $4 billion last year, but its growth forecast for 2026 has been downgraded, and its valuation has decreased from $42 billion to $35 billion.