Release2026-08-04

Black Forest Labs made the video generation part of its FLUX 3 model generally available through its own programming interface and selected partners. The model produces clips of up to 20 seconds with sound and dialogue created alongside the picture, starting from text, a still image, keyframes or a few seconds of existing footage. A cheaper draft mode returns a quick preview before a full-quality render. Image generation and an open-weight version are still to come.

What changed

FLUX 3's video generation had been announced but was not generally available.

What it unlocks

Generating short video clips with matching speech, sound effects and lip-sync in many languages from a text prompt, an image, keyframes or an existing clip.

  • clips up to 20 seconds
  • 720p HD, 1080p via upscaling
  • up to 4 seconds of video as input
  • ELO 1135 for text-to-video

What you need to act on it

  • BFL API access or a partner platform

Send this to someone who needs it

Shares the story and its sources. Nothing about you.

What does this mean for your job?

This is the story as everyone gets it. Once a week we send you the version written for your role — what changed, why it matters for the work you actually do, and one thing to try. Free while we tune it.