Stable Diffusion
In plain wordsA 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.
A family of text-to-image modelstext-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. released by Stability AI with downloadable weights. The original 2022 models are latent diffusion modelslatent 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.: a VAEvariational autoencoder, n.: A neural network that learns to squeeze data into a compact, slightly fuzzy code and rebuild it, arranged so that decoding random codes produces new, plausible examples. compresses 512 × 512 images to 64 × 64 × 4 latents, an 860-million-parameter U-NetU-Net, prop. n.: An image network shaped like the letter U: one half shrinks the picture to understand it, the other enlarges it back, with shortcuts between the halves so fine detail is not lost. denoises them, and a frozen CLIPCLIP, prop. n.: An OpenAI model that learned to match pictures with their captions from hundreds of millions of web examples, so it can judge how well any piece of text describes any image. ViT-L/14 text encoder (123 million parameters) supplies the prompt through cross-attentioncross-attention, n.: Attention between two different inputs: one sequence consults another, as when a translation model writing French looks back at the English sentence, or an image generator consults the text prompt.; training used subsets of LAION-5BLAION-5B, prop. n.: A free list of nearly six billion links to online images paired with their captions, released in 2022 and used to train image generators such as Stable Diffusion.. Successors include Stable Diffusion 2 (November 2022), SDXL (July 2023; a 3.5-billion-parameter base model with two text encoders), Stable Diffusion 3 (announced February 2024; a 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. with the MMDiT architecture) and Stable Diffusion 3.5 (October 2024; Large at 8 billion parameters and Medium at 2.5 billion), which as of October 2026 remained the latest release. Because the weights run on consumer GPUs, Stable Diffusion seeded a large open ecosystem — community web interfaces, LoRAlow-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. fine-tunes, ControlNetControlNet, prop. n.: An add-on for image generators that makes them follow a guide image — a sketch, a stick-figure pose or a depth map — so the layout comes out as intended., DreamBoothDreamBooth, prop. n.: A way to teach an image generator one specific subject — your dog, your face, a product — from a handful of photos, so it can draw that exact subject in new scenes. and textual inversiontextual inversion, n.: Teaching an image generator a new 'word' for a specific object or style from a few example pictures, leaving the model itself unchanged — only one vocabulary entry is added. — and popularized much of today's image-generation vocabulary.
We fine-tuned Stable Diffusion on our product photos with a LoRA overnight.
Origin
Developed by the CompVis group at LMU Munich and Runway, building on the latent diffusion paper of Robin Rombach and colleagues, with compute from Stability AI and training data from LAION. Version 1 was released publicly on 22 August 2022 under the CreativeML OpenRAIL-M license; later models use Stability AI's own licenses, including its Community License.
Usage note
Litigation over its training data includes Getty Images v. Stability AI, decided largely in Stability's favor by the High Court of England and Wales in November 2025, and the US artists' class action Andersen v. Stability AI. Downloadable weights under use restrictions make it an 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. rather than open-source AIopen-source AI, n.: AI released, ideally with its code and details of its training data and not just the finished model, so that anyone may use, inspect, alter, share and even rebuild it. How open is open enough is debated. in the strict sense.
Related
- 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.
- text-to-image modeltext-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.
- 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.
- LAION-5BLAION-5B, prop. n.: A free list of nearly six billion links to online images paired with their captions, released in 2022 and used to train image generators such as Stable Diffusion.
- ControlNetControlNet, prop. n.: An add-on for image generators that makes them follow a guide image — a sketch, a stick-figure pose or a depth map — so the layout comes out as intended.
- FLUXFLUX, prop. n.: A 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.
- AI copyright litigationAI copyright litigation, n.: The many lawsuits in which authors, artists, news organizations, record labels and others accuse AI companies of copying their work without permission to build or run AI models.
Sources
- Rombach, R., Blattmann, A., Lorenz, D., Esser, P. & Ommer, B. (2022). High-Resolution Image Synthesis with Latent Diffusion Models. CVPR 2022. arXiv:2112.10752. arxiv.org/abs/2112.10752
- Esser, P. et al. (2024). Scaling Rectified Flow Transformers for High-Resolution Image Synthesis. arXiv:2403.03206. arxiv.org/abs/2403.03206
Mentioned in
AI artAI art, n.: Images, music or other artworks made with the help of AI — from early computer drawing programs to today's prompt-based image generators — and the debate over whether and how such work counts as art., diffusion modeldiffusion model, n.: An AI that learns to create images, sound or video by practicing noise removal: shown pictures buried under static, it learns to clean them up, then builds new ones starting from pure static., MidjourneyMidjourney, prop. n.: A popular paid AI image generator, first used by typing prompts into the chat app Discord, known for its polished, painterly look; it can also animate images into short videos., text-to-image modeltext-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., CLIPCLIP, prop. n.: An OpenAI model that learned to match pictures with their captions from hundreds of millions of web examples, so it can judge how well any piece of text describes any image., ControlNetControlNet, prop. n.: An add-on for image generators that makes them follow a guide image — a sketch, a stick-figure pose or a depth map — so the layout comes out as intended., CVPRCVPR, prop. n.: The leading yearly conference on computer vision — teaching computers to understand images and video — and one of the most cited publication venues in any field of science., DALL·EDALL·E, prop. n.: OpenAI's pioneering series of image generators, named after the painter Salvador Dalí and the robot WALL·E, which showed in 2021–2023 that AI could draw almost anything described in words., DreamBoothDreamBooth, prop. n.: A way to teach an image generator one specific subject — your dog, your face, a product — from a handful of photos, so it can draw that exact subject in new scenes., FLUXFLUX, prop. n.: A 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., LAIONLAION, prop. n.: A mostly volunteer group, registered as a German nonprofit, that compiled free lists of billions of online pictures and their captions. Stable Diffusion and other image generators were trained on them., 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., model licensemodel license, n.: The legal terms under which an AI model may be downloaded and used — for example, whether businesses may use it, whether changes may be shared, and which uses are banned., negative promptnegative prompt, n.: A second prompt listing what you do not want in a generated image — 'blurry, extra fingers, watermark' — which the generator then steers away from., Stability AIStability AI, prop. n.: A London company that paid for and released Stable Diffusion, a free image-generating AI anyone could download, and later models for video, audio and 3D., U-NetU-Net, prop. n.: An image network shaped like the letter U: one half shrinks the picture to understand it, the other enlarges it back, with shortcuts between the halves so fine detail is not lost., Black Forest LabsBlack Forest Labs, prop. n.: A German AI company founded in 2024 by original creators of Stable Diffusion, which makes the FLUX image-generation models., LAION-5BLAION-5B, prop. n.: A free list of nearly six billion links to online images paired with their captions, released in 2022 and used to train image generators such as Stable Diffusion., 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., RunwayRunway, prop. n.: A New York company that makes AI tools for generating and editing video, used by filmmakers and artists; it helped create the image generator Stable Diffusion., textual inversiontextual inversion, n.: Teaching an image generator a new 'word' for a specific object or style from a few example pictures, leaving the model itself unchanged — only one vocabulary entry is added., VQGANVQGAN, n.: An image compressor that turns a picture into a small grid of codebook 'tiles' while keeping it sharp, used as the vocabulary of many image generators and of an early wave of AI art.