OpenAI is actively working on improving its Sora platform, testing image generation capabilities in addition to the existing video creation features. In particular, a new hidden toggle has appeared in the interface, allowing users to choose between generating videos and images directly in the input panel. When selecting images, users are prompted to describe the desired image, as well as choose the format, quality, and number of images to generate.
This feature is not yet available to the general public, but an Images Internal category has already appeared in the navigation panel on the left. Although it currently only opens videos, in the future users will likely be able to view an image feed. The exact image generation capabilities that will be available are still unknown, but it is likely that they will be powered by the sora-turbo model.
In addition, OpenAI is working on updating Sora’s video feed, which will be divided into “Best” and “Top” categories. The “Best” category will likely be similar to the current recommended feed, while “Top” may allow filtering by a specific period and possibly ranking videos by the number of likes or other criteria.
There is no official confirmation from OpenAI regarding the release of the new DALL-E 4 model yet, but the AI community is eagerly awaiting updates. Meanwhile, users hope that image generation features will be implemented soon, as the existing DALL-E 3 model is already somewhat outdated.
