FLUX image models

What it is

FLUX is a family of generative AI models that create and edit images from text descriptions, developed by Black Forest Labs. The company was founded in 2024 in Freiburg im Breisgau by Robin Rombach, Andreas Blattmann and Patrick Esser, researchers whose earlier work at LMU Munich led to Stable Diffusion. The first FLUX.1 models were announced in August 2024, and FLUX.2 followed in November 2025.

How it works

FLUX models are trained on large collections of images and descriptions. They learn to generate images by starting from random noise and refining it step by step towards a picture that matches the text prompt. The company releases models in different versions: some with open weights that anyone can download and run, some under non-commercial licences and some available only through a paid interface. FLUX.1 Kontext, released in 2025, added the ability to edit existing images with text instructions, and FLUX.2 uses a large vision-language model to understand prompts.

Why it stands out

Black Forest Labs is one of the few European companies operating at the frontier of image generation, a field dominated by large US companies. Releasing open models alongside commercial versions has made FLUX popular with developers and creative communities, while companies license the commercial models. The team's background in Stable Diffusion research shows the depth of German academic expertise in generative AI.

Where you meet it

FLUX models are integrated into image-editing tools, design platforms, chatbots and apps, often without users knowing which model is running underneath. Developers run the open versions on their own hardware, and creative agencies use them for concept art, advertising and product visualisation.