OpenAI launches Images 2.6 with GIF creation and faster speeds

New animation and speed features

OpenAI has officially released Images 2.6, the latest iteration of its image generation technology. This update is now available to users across ChatGPT, Codex, and the company's API. The primary goal of this release is to provide higher visual quality while significantly reducing the time it takes to create an image.

One of the most notable additions is the ability to create animated GIFs from still images. This functionality allows users to transform a static picture or even a basic sketch into a short animation. Such a feature is particularly useful for creating brand mascots, social media reactions, and various types of memes.

In addition to the new animation capabilities, the model is 50 per cent faster than the version that came before it. This speed boost is intended to make the creative process more fluid for users who need to iterate quickly. The update also aims to produce more professional-looking visuals, such as product images and posters.

Consistency and editing tools

A major hurdle in AI image generation has often been maintaining the same details across multiple changes. Images 2.6 is designed to solve this by keeping important elements consistent. This includes preserving specific character features, product designs, and established brand elements when a user requests an edit.

To help with these modifications, OpenAI has improved how the model handles image edits. Users can now maintain a sense of continuity even as they change various parts of a scene. This is a significant step forward for creators who need to refine a specific image without the AI changing every other detail in the frame.

The update also brings new tools for image markup. This feature allows users to point out specific areas of an image that require changes. While this was previously available in ChatGPT, it is now integrated into the broader Images 2.6 framework to provide a more cohesive editing experience for all users.

Specialized API models for developers

For developers using the OpenAI API, the company has introduced two distinct models tailored for different needs. These are named GPT-Image-2.6 Flare and GPT-Image-2.6 Sunburst. By splitting the model into two versions, developers can choose the specific balance of speed and quality that fits their unique application requirements.

The Flare model is built specifically for high-speed generation. It is capable of creating images in as little as six seconds when the quality setting is set to low. This makes it an ideal choice for applications where rapid output is more important than every minute visual detail. It serves as a high-velocity option for real-time needs.

Conversely, the Sunburst model is designed for more detailed image generation. It focuses on maintaining better consistency across multiple edits and producing high-fidelity results. This model is better suited for projects where the visual integrity of a character or product must remain perfectly stable throughout a series of modifications.

Technical controls and exports

OpenAI has also introduced a new complexity parameter for developers using the API. This tool allows for more precise control over how much planning the model performs before it begins generating an image. By adjusting this setting, developers can dictate the level of detail the AI attempts to include in the final output.

A lower complexity setting is particularly useful for quick edits where the model does not need to perform heavy computation. On the other hand, a higher setting helps the model handle more detailed and complex requests. This gives developers a way to manage the trade-off between processing power and the intricacy of the final image.

Furthermore, the update includes new SVG export support. SVG is a vector graphics format that allows designs to be scaled without losing quality. This means users can now take their AI-generated designs and use them across various platforms, including websites, mobile apps, and physical print materials.

Global availability and access

Images 2.6 is currently available globally for all users. It can be accessed through the web, desktop, and mobile versions of ChatGPT. This ensures that the new features, including GIF creation and the faster generation speeds, are reachable for a wide audience of both casual and professional users.

The update is included in all current subscription plans. This means that anyone with a standard subscription to ChatGPT can begin using the new model immediately. The company has made it a priority to ensure that these improvements are accessible to the existing user base without requiring additional tiers.

As reported by Digit on MSN, the release marks a significant step in the evolution of OpenAI's image tools. By offering both a fast

label__