MUMBAI: Still images are getting ready to move, and OpenAI wants users to do the animating. The company has introduced Images 2.6 across ChatGPT and its API, adding GIF creation, faster image generation and more precise editing tools as the use of AI-generated visuals continues to climb.
The update allows users to turn a text prompt or an uploaded photograph into a GIF directly within ChatGPT. They can also adjust animation speed, edit individual frames and share the finished creation, taking image generation beyond static pictures and into short-form animation.
The feature opens up a range of everyday creative possibilities. Users can create animated greetings, turn photographs into moving visuals or give familiar characters a fresh spin. OpenAI has highlighted festive applications such as animated Diwali greetings, cartoon-style portraits and personalised caricatures.
The process is designed to make animation more accessible without requiring users to work with separate editing software or build an animation frame by frame from scratch. Users can begin with an idea or an existing image, generate a moving version and then make adjustments to individual frames and playback speed.
Another addition is the ability to save and tag reference images of people, pets and characters. Users can retain these references and bring them into future image-generation requests without having to upload the same photographs repeatedly.
The feature is aimed at maintaining continuity across multiple creations. For instance, someone creating a series of illustrations featuring the same character or pet can reuse a saved reference rather than starting from a fresh upload each time.
This could be particularly useful for users creating recurring characters, personalised artwork or a series of visuals in which the subject’s appearance needs to remain consistent. It also makes the process less repetitive, as users do not have to keep supplying the same reference material for every new prompt.
The launch comes amid a sharp increase in the volume of images being generated across OpenAI’s platforms. The company said users now create more than 4 billion images every week across ChatGPT and its API.
That is up from more than 3 billion weekly images reported eight weeks earlier, when OpenAI launched Images 2.5. The increase highlights the growing use of generative AI for visual content, from casual experimentation and personalised artwork to more regular creative workflows.
The latest update also reflects a broader shift in how users interact with image-generation tools. Rather than simply producing an image from a written instruction, the experience is increasingly built around editing, reusing references and creating variations of existing visuals.
For developers, OpenAI is introducing two API models, GPT-Image-2.6 Flare and Sunburst, designed to deliver image generation up to 50 per cent faster.
Flare is geared towards speed, generating images in as little as six seconds at low quality. Sunburst, meanwhile, focuses on higher-quality outputs and greater consistency across edits at medium and high quality.
The two models cater to different requirements. Developers working on applications where response time is important can use Flare for quicker generation, while those prioritising visual quality and consistency across successive edits can turn to Sunburst.
The API update also brings image markup, a feature previously available in ChatGPT, allowing developers to make more targeted edits. Instead of relying entirely on broad text instructions, users can identify specific areas of an image that need changing.
OpenAI is also adding SVG export capabilities, giving developers an option to work with scalable vector graphics. Another addition is a complexity control, which determines how much planning the model undertakes when generating an image. This gives developers greater control over the balance between generation speed and the effort invested in producing an output.
Images 2.6 is available across ChatGPT, ChatGPT Work and Codex on the web, desktop and mobile, across all tiers. The GPT-Image-2.6 Flare and Sunburst models are also available through the API.
With GIF creation, reusable reference images, frame-level editing and faster developer models, the update expands the role of image generation from producing individual visuals to supporting more involved creative workflows. The bigger shift is that users are no longer just asking AI to draw something; they are increasingly asking it to animate, refine and keep building on the same idea.