ChatGPT Adds GIF Creation From Text and Photos

ChatGPT is gaining a long-awaited feature: animated GIF creation. Users will be able to turn a written prompt or an uploaded photo into a short moving image. The feature arrives as part of OpenAI's Images 2.6 update and is rolling out gradually.

GIFs tend to step in when words fall short in group chats, and that need has mostly been met by ready-made libraries such as Giphy and Tenor. ChatGPT's new feature takes a different route: instead of searching for existing content, users describe the animation they want, have the AI generate it and then edit the result. The speed and frame-level controls set it apart from a simple generation tool.

GIFs from text and from photos

According to the announcement, ChatGPT will accept two kinds of input for animation. The first is a plain text prompt. The second is an uploaded photo that the system converts into a moving image. Personal photos or a specific idea can therefore become a short animation.

That approach has the potential to produce more personalized results than the generic GIFs found in stock libraries. A reaction image for a group of friends or colleagues could be shaped around the user's own description and their own photo.

Image generation is already one of ChatGPT's widely used capabilities, and adding GIF support shows the tool moving beyond static images into moving content. The addition could also make it easier for ordinary users to put together short animations.

Speed control and frame-by-frame editing

One of the more notable aspects is the ability to edit a generated GIF. Users can adjust playback speed between 0.5x and 2x, slowing an animation down to add emphasis or speeding it up for a punchier effect.

Individual frames can also be edited. Rather than regenerating an animation from scratch when the first attempt isn't quite right, users can fine-tune specific frames. The range from half speed to double speed also means the same animation can be reused in different tones.

Frame editing could save time in cases where the output is mostly correct but wrong in the details. Users can focus on the problem frame instead of re-prompting the whole animation, which may shorten the trial-and-error process.

Images 2.6 and availability

GIF creation comes with OpenAI's Images 2.6 update. The update is being made available in Codex as well as ChatGPT and is rolling out gradually across web, desktop and mobile. Another element of the update is faster image generation.

Details on which users will get the feature and when go no further than the phrase gradual rollout, so some users may see GIF creation before others. More will become clear as additional details are shared; for now, what is known is limited to animation from text and photos, 0.5x to 2x speed adjustment and frame editing.