Black Forest Labs Launches FLUX 3 Video Model with Native Audio

German AI company Black Forest Labs announced FLUX 3, a video generation model that creates up to 20-second 1080p clips with native audio. It uses one architecture for images, video, and audio, supporting text-to-video, image-to-video, video continuation, keyframe-to-video, and multi-shot editing.

In benchmarks, FLUX 3 scored 1135 on text-to-video, beating Gemini Omni Flash and Minimax H3, and 1051 on image-to-video, ahead of Seedance 2.0 and Minimax H3. Pricing is per second: $0.06 for Draft HD, $0.17 for regular HD, and $0.29 for full HD, with higher rates for video-to-video.