Over the next few weeks and months, we will make the following capabilities available

Open-weight access to a multimodal backbone, for content creation (video, audio and image) and action prediction. (“FLUX 3 Dev”)

We will also release more technical details on the underlying approach.

- Frivolous use of the term World Model.

- Claims 20 seconds of video, shows only jumpcuts.

Coming soon!

But then again I heard the downers have always been the first to leave their dung comments here so let's see...

    - Video and audio generation and editing through APIs and private weight access. (“FLUX 3 Video”)
    - Action prediction through selected research and commercial partners, beginning with mimic robotics (“FLUX-mimic and FLUX 3 Action”)
    - Image synthesis and editing through APIs and private weight access. (“FLUX 3 Image”)
    - Open-weight access to a multimodal backbone, for content creation (video, audio and image) and action prediction. (“FLUX 3 Dev”)

a model must learn a representation of the world: […] and how events sound

I honestly hope they put an unrealistic amount of wilhelm scream into the learning process, just for fun.

Isn’t video + audio all you need?

It jointly learns from images, videos, and audio within a unified architecture, because what it needs to learn is not any one of these elements in isolation.

I'm confused, videos contain images and audio ...?

Of course there will be feuds from robots of different family groups but they will be minimal as it quickly becomes symmetrical robot conflict with high casualties as they learn too fast from each other, it's likely those will be avoided, it will be after all much easier to confront humans for any given resources.

Truly a pinnacle for technology, albeit perhaps not for mankind.

our mission to develop real-world visual intelligence

Visual is mono-modal, isn't it?