AI
ChatGPT adds native GIF generation and frame-by-frame editing
OpenAI introduced GIF generation with the Images 2.6 model, allowing users to create animations from text prompts or uploaded photos, adjust speed, edit individual frames, and share. The update covers ChatGPT and Codex on the web, desktop, and mobile, while also improving image generation speed.
The Verge reports that ChatGPT can now directly generate GIFs based on prompts or photos uploaded by users. Users can also adjust animation playback speed, view and edit frames one by one, and share the generated results with others.
Previously, ChatGPT could not natively generate GIFs in a single step. An example provided by OpenAI shows that after a user asks, “Can you make a GIF of a cat dancing?” the system can directly generate an animation and provide options to change playback speed and view it frame by frame.
This capability was introduced with OpenAI's new Images 2.6 model. OpenAI said Images 2.6 also speeds up image generation in ChatGPT; users currently generate more than 4 billion images per week through ChatGPT and the API, up from the 3 billion per week announced in September.
Images 2.6 is rolling out to all ChatGPT and Codex users on the web, desktop, and mobile apps. OpenAI also released GPT-Image-2.6 Flare and Sunburst in the API for faster, higher-quality image generation.