"ChatGPT" gets new tools for creating and editing animated images

OpenAI announced the launch of the image model "Images 2.6", which enables creating and editing animated GIF files within ChatGPT, with a 50% improvement in generation speed and new editing tools.

OpenAI announced on Wednesday the launch of the image model "Images 2.6", adding animated GIF creation, along with speed improvements and additional editing tools, as part of an update covering ChatGPT, Codex, and the API. The company said the new capabilities are becoming available depending on the platform and usage method, noting that the new model provides 50% faster image generation compared to the previous version, with improved preservation of character and product features when making repeated edits to an image. According to the company's announcement, the "Flare" version is offered as the fastest option, while the "Sunburst" version targets uses requiring a higher level of detail and consistency, especially when selecting medium or high quality levels. The Verge reported that users can now create an animated image from a text description or from an image they upload to the service, then adjust playback speed and work on individual frames within the same file. The report added that this update allows performing these steps directly within ChatGPT, instead of relying on a multi-step workflow as before. In the API, OpenAI added the ability to specify areas to be edited within the image, along with support for exporting graphics in SVG format. It also introduced a new setting to control the amount of planning the model allocates to executing the request, explaining that this setting is separate from the image quality option; the former relates to the complexity of the composition, while the latter determines the level of detail.