OpenAI has introduced Images 2.6, an upgraded image-generation model that brings faster image creation, animated GIF generation and improved editing features to ChatGPT. The update expands the platform’s visual capabilities, allowing users to create, animate and modify images through a single interface.
Announced on November 4, the latest version introduces several features aimed at making AI-powered visual creation more accessible. Users can now turn text prompts or uploaded photographs into animated GIFs directly within ChatGPT. They can also adjust animation speeds, review individual frames and make changes to specific parts of an animation.
The introduction of GIF generation marks an expansion of ChatGPT’s image-generation capabilities. Previously, users often needed separate editing applications to convert static images into animations. With Images 2.6, much of this process can now be completed within ChatGPT.
Users can describe the type of animation they want through a text prompt or upload an existing photograph and request a moving version. The model then generates an animation based on those instructions.
Another important addition is frame-by-frame editing. Instead of accepting the initial output, users can inspect individual frames and modify selected elements. They can also change the speed of an animation to achieve the desired visual effect.
These features could benefit social media creators, graphic designers and businesses producing digital content. Animated characters, reaction GIFs, promotional visuals and short storyboards are among the possible applications. The ability to edit individual frames also gives creators more flexibility when experimenting with visual concepts.
Businesses could use the new capabilities to develop animated advertisements, product demonstrations and branded content. Designers may also find the feature useful for testing character movements and visual ideas before investing in more elaborate video production.
OpenAI says Images 2.6 generates images 50% faster than its previous version. The improvement is expected to reduce waiting times and allow users to experiment with different prompts and editing instructions more efficiently.
Alongside the ChatGPT update, OpenAI has introduced two specialised versions of its image-generation model for developers using the API. These are GPT-Image-2.6 Flare and GPT-Image-2.6 Sunburst, which cater to different image-generation requirements.
Flare is designed for speed and rapid image creation. According to OpenAI, it can generate an image in as little as six seconds when the quality setting is kept low. This could make it suitable for applications that require frequent image generation, quick modifications and faster turnaround times.
Sunburst, meanwhile, focuses on visual detail, precision and consistency. It is intended for projects that require more sophisticated image generation and reliable results across multiple editing rounds, particularly at medium and high quality settings.
Maintaining consistency across repeated edits has been a challenge for AI image-generation systems. Details such as a character’s appearance, a product’s shape or a brand’s visual identity can change when an image undergoes several modifications. The latest model aims to make these editing processes more predictable and preserve important visual elements.
OpenAI has also extended its image markup feature to the API. Previously available in ChatGPT, the tool allows users to highlight specific areas of an image and indicate where modifications are required. This provides a more direct way to communicate editing instructions, reducing the need for lengthy text descriptions.
Another addition is support for exporting images in Scalable Vector Graphics (SVG) format. SVG images can be resized without losing quality, making them suitable for websites, applications, logos, icons and print designs. The feature could help developers and designers integrate AI-generated visuals into their existing workflows.
The update also gives developers access to a complexity setting that determines how much planning the model uses before generating an image. Lower settings are intended for simpler requests and routine edits, while medium and high settings can support more complicated scenes involving multiple objects and detailed visual instructions.
This allows developers to balance speed and image quality according to their requirements. Straightforward editing tasks may not need extensive planning, while complex visual projects could benefit from higher complexity settings.
The launch comes as AI-generated visual content becomes increasingly popular. OpenAI said users worldwide generate approximately four billion images every week across ChatGPT, Codex and its API. This amounts to more than 500 million images daily, highlighting the growing importance of image generation within its AI ecosystem.
Images 2.6 is available globally across ChatGPT and Codex on desktop, mobile and web, covering all subscription tiers. Developers can access GPT-Image-2.6 Flare and GPT-Image-2.6 Sunburst through the API.
With faster generation, GIF creation, frame-level editing and additional developer tools, OpenAI is expanding ChatGPT’s role in digital content production. The platform is moving beyond static image generation towards a more comprehensive creative environment, allowing users to generate, animate and refine visual content without switching between multiple applications.