Keep the clip going
Feed a finished video back in and Flux 3 continues it from the last frame, matching the motion and scene it inherits. Build a longer sequence a piece at a time instead of starting over.
Why creators choose Flux 3
Start CreatingFirst and last frame
Give the model a starting image and an ending image and it generates the movement between them, so the clip begins and ends exactly where you decided it should.
The widest framing
From 21:9 ultrawide through 2:1, square, and full vertical. Shoot for the format you are publishing to rather than cropping something down afterwards and losing half the frame.
From the FLUX team
Black Forest Labs’ first move into video, carrying over the prompt adherence the FLUX image models are known for. What you describe is what tends to come back.
How to create with Flux 3
Start creating with Black Forest Labs’ Flux 3 by following a few simple steps directly inside the DaVinci AI Toolkit.
Who Flux 3 is for
Flux 3 by Black Forest Labs suits creators who want to build a sequence in pieces rather than settle for whatever fits in one generation.
Content creators
Start with one clip and extend it until the story is actually told. Useful when a single generation runs out before the idea does.
Agencies and brands
Set the first and last frame around a product beat and let the model handle the movement between them. Extend the clip when the cut needs more room.
Filmmakers and creative teams
Build a shot in stages. Generate, review, extend from the last frame, and keep the scene continuous instead of stitching unrelated takes together.
Content creators
Start with one clip and extend it until the story is actually told. Useful when a single generation runs out before the idea does.
Agencies and brands
Set the first and last frame around a product beat and let the model handle the movement between them. Extend the clip when the cut needs more room.
Filmmakers and creative teams
Build a shot in stages. Generate, review, extend from the last frame, and keep the scene continuous instead of stitching unrelated takes together.
Try Flux 3
Enter a text prompt, upload an image, or continue a clip you already have, and generate video with sound in seconds.
Powerful features
Flux 3 brings together video extension, keyframe control, and native audio to produce video from text, images, or footage you already have.
Text, image, and video
Generate from a written prompt, animate a single still, or feed in an existing clip and carry it forward.
Video extension
Continue a clip beyond its final frame, with the new footage picking up the motion and the scene of the source rather than starting over.
First and last frame control
Define the opening and closing images and let the model generate the transition between them.
Native audio generation
Sound is created alongside the visuals as part of the same generation, so the clip comes back complete.
Text, image, and video
Generate from a written prompt, animate a single still, or feed in an existing clip and carry it forward.
Video extension
Continue a clip beyond its final frame, with the new footage picking up the motion and the scene of the source rather than starting over.
First and last frame control
Define the opening and closing images and let the model generate the transition between them.
Native audio generation
Sound is created alongside the visuals as part of the same generation, so the clip comes back complete.
