OpenAI has once again taken an important step in the evolution of its image-creation capabilities. After all the improvements we saw in ChatGPT Images 2.5, Images 2.6 arrives, a version that lets us generate images much faster and turn them into animated GIFs. According to the company, the system generates images 50% faster and preserves details much better during subsequent edits, but there’s more.
From a static image to an animation in seconds
The big new feature in Images 2.6 is GIF creation. We can start from a simple description or a sketch, as before, and ask for a character to wave at us, for a pet to move in a certain direction, or for an element in the image to simply come to life. ChatGPT will handle the animation while maintaining the appearance and consistency of the objects throughout the animation.
Until now, creating animated GIFs with ChatGPT required preparing several images and then assembling them into a sequence. Now we can describe the movement we want directly and let Images 2.6 give us the finished result.
Faster images and much more precise edits
The other big new feature in this version is speed. GPT-Image-2.6 Flare can generate low-resolution images in just six seconds. Sunburst, meanwhile, prioritizes detail by spending a bit more time generating them. Once we have the image, we can touch it up while keeping characters and products consistent across the different edits.
If we use Images 2.6 through the API, we’ll find complexity controls, support for annotations, and the option to export in SVG for web designs and applications. Like the Photoshop integration in ChatGPT, the improvements in this new model give us a much simpler and more intuitive image-creation system. Now available in ChatGPT and Codex for all plans, it’s time to let your imagination run wild.
