Black Forest Labs debuts FLUX 3, a joint image-video-audio-robotics model, in gated early access · cho.sh
Black Forest Labs debuts FLUX 3, a joint image-video-audio-robotics model, in gated early access
AI video model releases
Black Forest Labs debuts FLUX 3, a joint image-video-audio-robotics model, in gated early access
Black Forest Labs launched FLUX 3, a single model jointly trained to generate images and up to 20-second audio/video clips from one prompt, and to extend to robotic vision and action control.
FLUX 3 ships as four product lines: FLUX 3 Video, FLUX 3 Image, FLUX 3 Action, and an upcoming open source FLUX 3 Dev; Video and Action enter a gated Early Access program anyone can apply to but BFL must approve.
No pricing, API access, production SLAs, or benchmark methodology have been announced; FLUX 3 Image is promised in coming weeks and FLUX 3 Dev's open weights are slated for later this year, unlike prior all-image FLUX Dev releases.
In preliminary internal preference tests on 10-second 720p text-to-video with audio, an early FLUX 3 candidate was preferred over Luma Ray 3.2 in 93% of comparisons, Runway Gen-4.5 in 77%, Grok Imagine Video in 69%, Kling v3 Pro in 60%, and tied at 52% against Seedance 2.0 and Google Gemini Omni Flash.
Seedance 2.0, ByteDance's model, is unavailable to most Western enterprises after Netflix, Warner Bros., Disney, Paramount, and Sony sent legal threats over alleged copyright infringement, prompting an indefinite pause on its international rollout.
cho.sh