Black Forest Labs Unveils FLUX.1, Redefining Text-to-Image AI

The new open-source model suite outperforms competitors in image detail, prompt adherence, and style diversity.

FLUX.1 features a hybrid architecture with 12 billion parameters, enhancing performance and efficiency.
Available in three variants—pro, dev, and schnell—catering to different user needs and applications.
Surpasses popular models like Midjourney v6.0 and DALL·E 3 in key areas, including human anatomy rendering.
Released under the Apache 2.0 license, promoting versatile use across personal, scientific, and commercial projects.
Black Forest Labs aims to expand into video generation, building on FLUX.1's foundation.