OpenAI unveiled ChatGPT Images 2.5 on Tuesday, bringing enhanced image generation capabilities and a fresh approach to image creation. The announcement includes the introduction of Sketch, a feature that allows users to input rough drawings directly into ChatGPT and receive refined, detailed images based on their sketches and instructions.
The Sketch feature functions by typing @Sketch into the chat interface, which opens a drawing window for users. Early testing demonstrated the capability to transform basic doodles, such as a simple mouse-drawn cat sketch, into photorealistic images according to user specifications. Additionally, the platform now supports direct commenting on specific areas within generated images, enabling targeted editing requests such as adjusting colors or other visual elements.
The updated ChatGPT Images 2.5 model incorporates several technical improvements. OpenAI reports that the new version produces images with enhanced natural lighting and textural detail compared to previous iterations. The model has also improved its ability to follow editing instructions across multiple interaction turns, providing users with greater control over iterative refinement. Image generation speed has been notably accelerated, with latency reductions of up to 50% compared to Images 2.0.
The feature rollout applies across multiple platforms and user tiers. ChatGPT Images 2.5 is now available to ChatGPT, ChatGPT Work, and Codex users on desktop, mobile, and web platforms, according to the company’s announcement.
Article Attribution | Read More at Article Source
Article summary produced by Claude AI