B.AI Launches AI-Powered Image Generation, First Phase Integrates GPT-Image-2

Odaily News — On September 16, B.AI officially launched its AI-powered image generation feature, with the first phase deeply integrating the GPT-Image-2 model. Available simultaneously on both web and API, it brings users and developers a creation experience with pixel-level control. The feature supports both text-to-image and image-to-image modes, allowing users to input text descriptions or reference images to precisely recreate visual ideas, and covers diverse styles including 3D, comics, photographic-grade portraits, and ink-wash painting. Users can freely adjust between 1K, 2K HD, and ultra-HD 4K resolutions and quality tiers, with support for multiple aspect ratios including 16:9, 3:2, 1:1, 2:3, and 9:16, as well as smart sizing. Up to 8 images can be generated simultaneously in a single session, accommodating both individual creative needs and developer integration requirements. Whether for quick personal inspiration sketches or large-scale team content production, B.AI's AI image generation enables every idea to be brought to life quickly with lower barriers and higher efficiency.