FLUX
In plain wordsA family of image-generating AI models from the German start-up Black Forest Labs, founded by researchers behind Stable Diffusion; some versions can be downloaded and run freely.
A family of text-to-imagetext-to-image model, n.: An AI system that creates a picture from a written description, such as 'a fox reading a newspaper, in watercolor', usually within seconds. and image-editing models from Black Forest Labs. FLUX.1 (August 2024) is a 12-billion-parameter rectified flowrectified flow, n.: A generator trained to move random noise toward real data along straight lines, which can then be straightened further so that a picture comes out in one or a few steps. transformer — a diffusion transformerdiffusion transformer, n.: A diffusion image or video generator whose core network is a transformer — the architecture behind chatbots — instead of the older image-specific design, which lets it improve more predictably as it grows. that processes text and image tokens jointly — released in three tiers: an API-only 'pro' model, an open-weight 'dev' model under a non-commercial license, and a step-distilleddiffusion distillation, n.: Training a fast image generator to imitate a slow diffusion model, so that a picture that once took fifty steps can be made in one to four. 'schnell' model under the Apache 2.0 license. Its downloadable versions became a widely used base for community fine-tunes and LoRAslow-rank adaptation, n.: A cheap way to customize a big AI model: instead of changing all its billions of numbers, you train a small add-on of a few million numbers that slightly adjusts how it behaves.. FLUX.1 Kontext (May 2025) added in-context instruction-based image editinginstruction-based image editing, n.: Changing a picture by telling an AI what to change — 'make it night', 'remove the car', 'give her a red coat' — instead of using editing tools by hand. from reference images, and the FLUX.2 series (November 2025) uses a 24-billion-parameter Mistral vision-language model to encode prompts. As of October 2026 its newest line was FLUX 3 (mid-2026), a multimodal family covering image, video and audio generation and action prediction for robots. FLUX models have been integrated into products including xAI's Grok (August–December 2024) and Adobe Photoshop's generative fill (2025).
Origin
Black Forest Labs, based in Freiburg im Breisgau, Germany, was founded in 2024 by Robin Rombach, Andreas Blattmann and Patrick Esser — co-authors of the latent diffusion and Stable Diffusion 3 papers who had left Stability AI — and launched FLUX.1 on 1 August 2024.
Usage note
Written FLUX by its makers and often 'Flux' in the press. 'Schnell' is German for 'fast'.
Related
- Stable DiffusionStable Diffusion, prop. n.: A widely used image-generating AI whose model files were released for anyone to download, so people could run it on their own computers and adapt it.
- rectified flowrectified flow, n.: A generator trained to move random noise toward real data along straight lines, which can then be straightened further so that a picture comes out in one or a few steps.
- diffusion transformerdiffusion transformer, n.: A diffusion image or video generator whose core network is a transformer — the architecture behind chatbots — instead of the older image-specific design, which lets it improve more predictably as it grows.
- latent diffusion modellatent diffusion model, n.: A diffusion image generator that works on a shrunken, compressed version of each picture instead of the full-size pixels, which makes it far cheaper to train and run.
- instruction-based image editinginstruction-based image editing, n.: Changing a picture by telling an AI what to change — 'make it night', 'remove the car', 'give her a red coat' — instead of using editing tools by hand.
- open-weight modelopen-weight model, n.: An AI model whose trained internal numbers are published for anyone to download and run on their own computers, even if the data and code used to make it stay private.
Sources
- Black Forest Labs: Batifol, S. et al. (2025). FLUX.1 Kontext: Flow Matching for In-Context Image Generation and Editing in Latent Space. arXiv:2506.15742. arxiv.org/abs/2506.15742
Mentioned in
Stable DiffusionStable Diffusion, prop. n.: A widely used image-generating AI whose model files were released for anyone to download, so people could run it on their own computers and adapt it., diffusion transformerdiffusion transformer, n.: A diffusion image or video generator whose core network is a transformer — the architecture behind chatbots — instead of the older image-specific design, which lets it improve more predictably as it grows., flow matchingflow matching, n.: A way to train generators that learns, for every point between random noise and real data, which direction to move, so that noise can be steered smoothly into new examples., instruction-based image editinginstruction-based image editing, n.: Changing a picture by telling an AI what to change — 'make it night', 'remove the car', 'give her a red coat' — instead of using editing tools by hand., latent diffusion modellatent diffusion model, n.: A diffusion image generator that works on a shrunken, compressed version of each picture instead of the full-size pixels, which makes it far cheaper to train and run., rectified flowrectified flow, n.: A generator trained to move random noise toward real data along straight lines, which can then be straightened further so that a picture comes out in one or a few steps.