AI Image
Canva releases its first image generation model
TechCrunch • • general
Key Points:
- Canva has launched its first image generation model, Canva Craft, designed to compete with similar models from Google, OpenAI, ByteDance, and Meta, featuring a two-stage process for detailed and customizable image creation.
- The model interprets user prompts to plan spatial relationships and text generation before producing images up to 2K resolution in up to 30 aspect ratios, enabling accurate object details and legible text with precise editing capabilities.
- Canva Craft was trained on human-designed templates from Canva’s library rather than photographs, allowing generated images to be converted into layered designs using Canva Magic, the company’s image-to-design transformer.
- The model is accessible to all Canva users via the Dream Lab tab and through the Canva API, with plans to integrate image generation into third-party chatbots like ChatGPT and Gemini via Canva’s MCP.
- Canva is developing two more cost-efficient versions, Craftfast and Craft-ultrafast, and has benefited from its 2024 acquisition of Leonardo.ai, whose researchers contributed to the new model’s development.
