OpenAI's Images 2.6 model adds animated GIFs to ChatGPT
Users can make GIFs from text or photos and edit them frame by frame. Developers get two new API image models.
OpenAI is rolling out a new image model, Images 2.6, and with it the ability to make animated GIFs in ChatGPT. Users can turn a written prompt or an uploaded photo into a GIF, then change its playback speed, view and edit individual frames, and share it. Until now, ChatGPT could not make a GIF in a single step. OpenAI says Images 2.6 also speeds up image generation. The model is rolling out to all ChatGPT and Codex users on the web, desktop and mobile. Developers get two new API models: GPT-Image-2.6 Flare, built for speed, and GPT-Image-2.6 Sunburst, built for quality. This follows the speed-and-quality split OpenAI introduced with Images 2.5 in September. OpenAI says people now make more than 4 billion images a week across ChatGPT and its API, up from 3 billion in September. These are the company's own figures and cover all image use, not only GIFs. According to TNW, the feature arrives months after OpenAI shut down Sora, its AI video app, in March. Short animation inside the chatbot gives users a limited form of motion generation without a separate video product.