The key facts

  • A GIF is created in one step from text or an uploaded image; frame and speed editing happens right in the chat
  • Images 2.6 is available worldwide on all ChatGPT and Codex plans
  • ChatGPT and API users create 4 billion images per week — a third more than in September

What has changed

ChatGPT has gained the ability to create GIF animations from a text prompt or based on an uploaded image. The feature is built into the new version of the image generation model, Images 2.6, as reported by The Verge. Here the sources differ on details: the Uzbek-language version of Spot.uz writes that previously ChatGPT could not create GIF files in a single step at all, while the Russian-language version clarifies that GIFs could be created, but it required several prompts. Both outlets agree on the main point: animation can now be created in one step.

What you can do with the result

After generation, the finished animation can be edited right in the chat: change the playback speed, view and fix individual frames, and share the result with other users. In its example, OpenAI shows the chatbot creating a GIF of a dancing cat and immediately providing adjustment tools.

Availability and load

Images 2.6 is already available worldwide in ChatGPT and Codex — on the web versions, desktop and mobile apps, across all subscription plans. The update is being rolled out gradually. According to OpenAI, ChatGPT and API users create more than 4 billion images per week — a third more than at the time of Images 2.5's release in September, when the figure was 3 billion.

Versions for developers

For developers working through the API, OpenAI has released two additional versions of the model: GPT-Image-2.6 Sunburst delivers higher image quality, while GPT-Image-2.6 Flare generates an image in 6 seconds.

chatgpt, openai