Black Forest Labs has opened early access to FLUX 3, its next major model family and the first FLUX release aimed beyond still-image generation.

The company describes FLUX 3 as a multimodal foundation model that learns from images, video and audio in one architecture. Its video system can generate clips up to 20 seconds long with native audio, including text-to-video, image-to-video and video-to-video modes. The company also says FLUX 3 will support image synthesis and editing through APIs and private weight access.

The more unusual part is robotics. Black Forest Labs says mimic robotics was an early partner on FLUX-mimic, a video-action model that combines the FLUX 3 backbone with robot-learning work for dexterous manipulation. The company says that effort is being tested on production tasks at Audi, tying the release to physical AI rather than only creative media tools.

Availability is still limited. Black Forest Labs is taking early-access requests now, while the public model page lists FLUX 3 as "coming soon." The launch plan separates the family into video, action, image and developer tracks. Video and action access begin with APIs and partners, image access is planned through APIs and private weights, and an open-weight FLUX 3 Dev backbone is planned later for content creation and action prediction.

That staged rollout makes the announcement less of a broad public launch than a signal of where frontier image labs are heading: shared models that generate media, understand motion and feed robot-control systems from the same visual foundation.