FLUX.1 [schnell] is a modified flow transformer with 1.2 billion parameters that generates images from text descriptions. It is renowned for its state-of-the-art output quality and competitive prompt-following capabilities, matching the performance of closed-source alternatives. The model is trained using latent adversarial diffusion distillation, allowing for the generation of high-quality images in as few as one to four steps. FLUX.1 [schnell] is released under the Apache-2.0 license, making it suitable for personal, scientific, and commercial use.