Stable Diffusion Creators Raise $31 Million for New FLUX.1 Image Model
The original creators of Stable Diffusion launch Black Forest Labs and unveil the powerful FLUX.1 AI image suite to rival closed-source giants. The new startup secures $31 million in seed funding to advance open-source generative AI.
Black Forest Labs, a new startup founded by the original creators of Stable Diffusion, unveils its FLUX.1 text-to-image model suite and secures $31 million in seed funding. Andreessen Horowitz leads the investment round, with additional backing from prominent tech figures. The company focuses on developing cutting-edge generative deep learning models for images and videos while prioritizing transparency and widespread accessibility.
The FLUX.1 suite offers three distinct variants to serve different user needs, including a closed-source pro API, a non-commercial open-weight dev version, and a fast Apache 2.0 licensed model for local development. All three models feature an impressive 12 billion parameters and use a hybrid architecture of multimodal and parallel diffusion transformer blocks. Early demonstrations show that the image quality rivals or even surpasses popular closed-source systems like Midjourney v6.0 and DALL-E 3.
This launch injects fresh energy into the open-source AI community following recent instability at Stability AI. Alongside the model release, Black Forest Labs establishes strict usage guidelines that prohibit the creation of false information, non-consensual imagery, and other harmful content. By combining top-tier performance with responsible development practices, the startup aims to accelerate innovation across fields like graphic design and scientific visualization.