Flux 3 Max

Sharp, prompt-obedient AI video from text, images or keyframes. 5 to 20 seconds with native sound.

FLUX 3 Max: AI Video That Follows Your Prompt, Fast

FLUX 3 Max is the sharpest, most obedient version of Black Forest Labs’ FLUX 3 video model, and you can use it on Mitte. It is tuned to do what you actually wrote: the actions in the order you describe them, the camera move you asked for, the light and mood you set. And it is quick. A 10-second clip comes back in about the time it takes to watch it.

Every video has sound built in: ambience, sound effects and spoken dialogue, generated together with the picture so they land on the right moment. Start from a prompt, bring a still image to life, or pin your own images to exact moments and let FLUX 3 Max fill in the motion between them.

What can you do with it?

How to use it

  1. Pick a mode. Text to Video, First / Last Frame, or Keyframes.
  2. Add your images. A starting image, a start and end pair, or a set of keyframes. Nothing at all for Text to Video.
  3. Write the shot. What happens in order, how the camera moves, the light, and what you hear. Put spoken lines in quotes.
  4. Set length, shape and quality. 5 to 20 seconds, seven aspect ratios from cinematic 21:9 to vertical 9:16, at 720p or 1080p.
  5. Generate. Download your MP4, share it, or keep building on it in Mitte.

What you get

One MP4 video per run, 5 to 20 seconds long, at 720p or 1080p, with native audio unless you switch it off. When you start from an image, the video keeps its shape.

Tips

All AI models & tools

See pricing and start creating on Mitte