[DigitalToday reporter Jinju Hong] Black Forest Labs has released its video-generation AI, FLUX 3 Video, to the public. The model can generate up to 20-second 1080p videos with audio.

Online media outlet Gigazine reported on Tuesday that the newly released FLUX 3 Video is the video-generation component of the multimodal model FLUX 3, which Black Forest Labs announced in July. FLUX 3 was introduced as a structure supporting image, video, audio and robot-control generation. At the time, it was available only to a limited number of users. With this release, video generation has been expanded to general users.

The focus is output quality and how it generates content. FLUX 3 Video supports 720p and 1080p resolution and can produce videos that include audio. It generates videos up to 20 seconds long. It offers text-to-video and a feature that extends video from an image input. Users can set the starting and ending frames and specify text to include in the video.

It also includes features aimed at reducing workflow. Black Forest Labs added a draft mode that lets users quickly check a rough result before producing a final output. Users review a draft first and then move to final generation.

On performance, it highlighted results from comparisons based on human evaluation. Black Forest Labs said FLUX 3 Video was rated the highest-performing model in both text-to-video and image-to-video generation. The article did not provide additional details on the benchmark conditions, but the company presented both generation methods as key strengths.

Its pricing is a per-second charging model based on its API. A 720p draft costs $0.06 per second, 720p costs $0.17 per second and 1080p costs $0.29 per second. Pricing varies by resolution and generation stage. That appears to target developer demand for separating fast concept checks from final output.

Black Forest Labs also released a follow-up roadmap. The company said it plans to improve control in FLUX 3 Video and add image, video and audio reference features. It also signalled development of FLUX 3 Image for image generation and editing, and said it would release FLUX 3 Dev, an open-weights version.

It also presented plans to expand resolution. Black Forest Labs said it is preparing 2K and 4K support for FLUX 3 Video and plans to release the model within days. By laying out high-resolution and open-model plans immediately after the public release, it appears to be placing weight on a strategy that combines a commercial API with public models.

Against that backdrop, FLUX 3 Video appears closer to a starting point for separating and expanding a video-generation AI product line than a simple feature addition. After releasing the multimodal FLUX 3, Black Forest Labs first made the video-generation component publicly available and laid out a schedule to roll out an image generation and editing model and an open-weights version. The market is expected to watch whether stronger control features, support for reference inputs and 2K and 4K compatibility translate into service competitiveness.

FLUX 3 Video is here. Serious, fun, creative, real, cinematic, whatever you need it to be. Native audio, Text to Video, Image to Video with multiple frames, video continuation, dialogue in multiple languages. Comes with Draft mode so you can explore ideas fast at a fraction of… pic.twitter.com/VPT8diaF4j