Download statistics from HuggingFace

Black Forest Labs, Formed by Stable Diffusion Team, Launches FLUX.1 Text-to-Image Models with High ELO Scores

Aug 1, 2024, 02:47 PM

Black Forest Labs, a new company formed by the original creators of the Stable Diffusion models, has announced the launch of its FLUX.1 suite of text-to-image models. The company, led by Robin Rombach, Patrick Esser, and Andreas Blattmann, aims to develop state-of-the-art generative deep learning models for media, pushing the boundaries of creativity, efficiency, and diversity. The FLUX.1 suite includes the FLUX.1[pro] model, available behind an API, and the FLUX.1[dev] model, a 12B parameter model available under non-commercial and Apache 2.0 licenses. Black Forest Labs claims that its models have achieved the strongest ELO scores and will soon release text-to-video models. The model weights are available on HuggingFace. The company has also partnered with venture capital firm Andreessen Horowitz (a16z), which has made a seed investment to support the development of these advanced visual AI models.