Generate and edit stills
BFL says FLUX 3 Image will handle complex prompts, multilingual text, references, varied styles, aspect ratios, and resolutions.
Announced July 23, 2026 · Independent launch tracker
One prompt.
Every dimension.
FLUX 3 is Black Forest Labs’ unified model for images, video, native audio, and action. We are building the direct, browser-based way to create with it when public access opens.
Be first when real generation opens.
Product launch only. No weekly newsletter.
01Image
02Video
03Audio
FLUX3WORLD MODEL
CREATEImage · video · sound
ACTPredict · respond
One architecture learns how the world looks, moves, sounds, and responds.
20 secvideo + native audio
1 modeljoint multimodal training
2026open weights planned
Facts checkedRead BFL’s announcement ↗
What FLUX 3 changes
Media are not separate tools anymore.
FLUX 3 learns the connections between a scene, its motion, its sound, and what can happen next. That shared understanding is the product—not a bundle of disconnected generators.
Create motion that sounds right
Generate up to 20-second clips with native audio from text, a starting frame, reference images, keyframes, or an existing video.
Predict what happens next
The shared multimodal backbone also supports action prediction, connecting generated worlds with robotics and physical AI.
Availability, without the hype
What you can access today.
Last verifiedJuly 24, 2026
ProductNowWhat is announced
FLUX 3 VideoEarly accessAPI and private weights
FLUX 3 ImageNot publicEarly access in the coming weeks
FLUX 3 DevPlannedOpen weights later in 2026
FLUX 3 ActionPartner accessSelected research and commercial rollout
Public pricing, API model IDs, rate limits, exact image resolutions, and FLUX 3 Dev licensing have not been announced. We will not invent them.
What we are building
From an idea to a living scene.
The product will open at the simplest useful workflow: describe what you need, add a reference when it matters, choose still or motion, and generate. Advanced controls will appear only when they help the result.
- PromptDescribe the scene in natural language.
- ReferenceKeep a subject, product, or visual direction consistent.
- CreateGenerate an image or a video with synchronized sound.
- RefineContinue, edit, or connect shots without rebuilding context.
Clear answers
FLUX 3 FAQ
What is FLUX 3?
FLUX 3 is Black Forest Labs’ multimodal foundation model for image, video, audio, and action prediction. Unlike separate models stitched together for each medium, it is trained across the modalities within one architecture.
Is FLUX 3 available now?
Partly. FLUX 3 Video and FLUX 3 Action are in controlled early access as of July 24, 2026. FLUX 3 Image is expected to enter early access in the coming weeks. General public access has not been announced.
Can I use the FLUX 3 image generator here?
Not yet. This independent site is preparing a browser-based generator for public FLUX 3 access. Join the launch list and we will notify you when real generation is available.
Does FLUX 3 generate video with audio?
Yes. Black Forest Labs says FLUX 3 can jointly generate video and native audio, including dialogue and sounds tied to visible events. Individual generations can be up to 20 seconds during the current early-access stage.
Will FLUX 3 be open source?
BFL has announced FLUX 3 Dev as an open-weight multimodal backbone planned for later in 2026. The exact license, model size, hardware requirements, and release date are not public yet.
Is this the official Black Forest Labs website?
No. flux-3.online is an independent product and is not affiliated with, endorsed by, or operated by Black Forest Labs. FLUX is a trademark of its respective owner.
The model is early. Your place does not have to be.
Get the launch signal.
We will email once the independent FLUX 3 generator is genuinely ready to use.