News for hackers. Stuff that matters.
Re:Flux 3 Image Klein: open weights, 5.6B parameters by badsectoracula on Friday October 16, 2026 @10:01PM (#50127140)
Looks like the people over /r/SD seem to like MOSS (i'm not up to date on the image stuff) but what i notice is that it is Apache2 licensed when every other openweight model seems to have some weird license (from a quick check Qwen-Image even downright forbids any commercial use at all).
Re:Flux 3 Image Klein: open weights, 5.6B parameters by vunderba on Friday October 16, 2026 @10:07PM (#50127184)
Yeah, unfortunately, a lot of them have been moving to their own bespoke, non-commercial licenses (Ideogram, Krea, Flux, Qwen 2, etc).

Annoyingly, the original Qwen-Image was actually Apache 2 before they switched over unfortunately.

In addition to Moss, Boogu Image 0.1 is a unified model which came out in June and is kind of being slept on right now, is actually pretty decent and falls under the Apache 2.0 license. It scored surprisingly high on my GenAI benchmark (comparable to Krea 2 and Ideogram 4).

https://github.com/boogu-project/Boogu-Image

Re:Flux 3 Image Klein: open weights, 5.6B parameters by badsectoracula on Saturday October 17, 2026 @12:00AM (#50127799)
Looks like MOSS was made by some Chinese university as part of an AI education/research program so i guess they don't have to care about funding as much as the other labs :-P