ChatGPT Adds GIF Maker With Images 2.6

OpenAI is gradually rolling out a GIF maker inside ChatGPT that can turn text prompts or uploaded photos into animated images, according to The Verge. The feature reportedly uses Images 2.6, an OpenAI image-generation model that has not received a standalone official announcement. Most users could not access the GIF maker during Decrypt’s testing, and OpenAI has not given a date for when all accounts will receive it.

Users can describe an animation in plain English or upload a photo and ask ChatGPT to animate it. After generation, they can adjust playback speed, inspect individual frames, edit the result and share the finished GIF. OpenAI demonstrated the feature with a request for a dancing cat. Decrypt’s test accounts could not generate GIFs; in its interaction, ChatGPT produced a sequence of frames rather than an animation.

The reported GIF feature is part of a broader Images 2.6 update that The Verge says is coming to ChatGPT and Codex on web, desktop and mobile. Developers are also getting two API models: GPT-Image-2.6 Flare, designed for faster generation, and GPT-Image-2.6 Sunburst, designed for higher-quality results. OpenAI has not published a standalone announcement describing Images 2.6’s capabilities, and Decrypt said it could not independently test the model’s performance.

The update follows the September 8 release of ChatGPT Images 2.5, which OpenAI said offered sharper details, richer textures, more natural lighting and more precise editing. OpenAI said image-generation latency was up to 50% lower than with Images 2.0. Its API models GPT-Image-2.5 Flare and GPT-Image-2.5 Sunburst were described as a fast default and a model for premium editing precision, respectively. In earlier testing across six categories, Images 2.5 and Google’s Nano Banana 2 each won three; Images 2.5 performed particularly well on illustrations but still made errors in rendered text.

The GIF rollout follows OpenAI’s shutdown of its standalone Sora video-generation app in March. The article describes GIFs as simpler than Sora’s videos, while noting that the feature puts motion generation inside ChatGPT without a separate application. OpenAI has not specified when GIF generation will reach every ChatGPT account.

OpenAI is gradually rolling out a GIF maker inside ChatGPT that can turn text prompts or uploaded photos into animated images, according to The Verge. The feature reportedly uses Images 2.6, an OpenAI image-generation model that has not received a standalone official announcement. Most users could not access the GIF maker during Decrypt’s testing, and OpenAI has not given a date for when all accounts will receive it.

Users can describe an animation in plain English or upload a photo and ask ChatGPT to animate it. After generation, they can adjust playback speed, inspect individual frames, edit the result and share the finished GIF. OpenAI demonstrated the feature with a request for a dancing cat. Decrypt’s test accounts could not generate GIFs; in its interaction, ChatGPT produced a sequence of frames rather than an animation.

The reported GIF feature is part of a broader Images 2.6 update that The Verge says is coming to ChatGPT and Codex on web, desktop and mobile. Developers are also getting two API models: GPT-Image-2.6 Flare, designed for faster generation, and GPT-Image-2.6 Sunburst, designed for higher-quality results. OpenAI has not published a standalone announcement describing Images 2.6’s capabilities, and Decrypt said it could not independently test the model’s performance.

The update follows the September 8 release of ChatGPT Images 2.5, which OpenAI said offered sharper details, richer textures, more natural lighting and more precise editing. OpenAI said image-generation latency was up to 50% lower than with Images 2.0. Its API models GPT-Image-2.5 Flare and GPT-Image-2.5 Sunburst were described as a fast default and a model for premium editing precision, respectively. In earlier testing across six categories, Images 2.5 and Google’s Nano Banana 2 each won three; Images 2.5 performed particularly well on illustrations but still made errors in rendered text.

The GIF rollout follows OpenAI’s shutdown of its standalone Sora video-generation app in March. The article describes GIFs as simpler than Sora’s videos, while noting that the feature puts motion generation inside ChatGPT without a separate application. OpenAI has not specified when GIF generation will reach every ChatGPT account.