FLUX 3 brings image, video, and audio into one model
Black Forest Labs is training FLUX 3 across images, video, and audio in one multimodal system. Video is in limited Early Access at BFL; access here, public pricing, and a general release date have not been announced.
By joining, you agree to receive one FLUX 3 availability email at your account address. This does not subscribe you to marketing updates.
Joint image, video, and audio training
What Black Forest Labs has confirmed
FLUX 3 mixes modalities instead of treating image, video, and sound as separate tools. The current announcement covers generation, editing, world understanding, and future action prediction.
Video with native audio
Generate video and audio together from text, an image, or a reference video. Early Access demonstrations cover clips up to 20 seconds, continuation, keyframes, and multilingual dialogue.
Image generation and editing
BFL says FLUX 3 can synthesize and edit images across styles, aspect ratios, and resolutions, with stronger complex-prompt handling and multilingual typography.
Audio tied to the scene
Audio is generated with video rather than added as a separate pass, allowing dialogue, ambient sound, and physical events to stay connected to the visual sequence.
Action prediction
The same backbone is being developed for action prediction and specialized robotics models. Broader access and technical specifications are still pending.
Availability is rolling out in stages
FLUX 3 Video is available through BFL Early Access. Image Early Access is planned for the following weeks, while Video, Action, Image, and Dev releases will move through separate early-access phases.
BFL has not published general-availability dates, public API pricing, or complete technical specifications. This page will only mark the model live when it is ready for reliable use here.
Create with available models today
You do not need to wait to build image and video workflows. The tools below are available now and use models with published pricing.
FLUX 3 questions
Is FLUX 3 available here?
Not yet. FLUX 3 Video is in limited Early Access through Black Forest Labs, but it is not ready for reliable use in this generator.
What can FLUX 3 generate?
BFL has confirmed image generation and editing, video with native audio, text-to-video, image-to-video, video-to-video, continuation, keyframes, multilingual dialogue, and work on action prediction.
When will FLUX 3 launch and how much will it cost?
General release dates and public API pricing have not been announced. The model family is rolling out through separate early-access phases.
What happens after I join the waitlist?
We record your interest once for your account and will send one email to your account address when FLUX 3 becomes available here.