Stable Diffusion web UI

AUTOMATIC1111/stable-diffusion-webuiReviewed by AI

What it does

A local web page for Stable Diffusion: you type a description and it makes images, repaints parts of a picture (inpainting and outpainting), enlarges photos and repairs faces, with many community extensions. It is for hobbyists and artists who want AI images without a subscription. Everything runs on your own PC and a capable graphics card helps a lot (NVIDIA recommended; AMD, Intel and Apple chips have separate guides). On Windows you install Python 3.10.6 and git, then run webui-user.bat. Models take several GB of disk.

How easy it is to install

C · install 1 or 2 tools · Windows

Windows C
macOS D
Linux C

Easiest on Windows: install Python 3.10.6 and git, then double-click webui-user.bat (or unzip the older NVIDIA-only sd.webui.zip and run update.bat and run.bat). Linux is similar but you first install system packages with apt, dnf, zypper or pacman. On a Mac the README sends you to a separate Apple Silicon guide with more terminal steps, and there is no ready installer anywhere.

Numbers on GitHub

Stars: 165,207on Oct 31, 2026
Downloads: 5,913,56314,580 this week
Growth: +53stars in 5 days

Stars over time

Facts

Runs on: Windows, macOS, Linux
You need to install: Python, Graphics card, Other tools
To build from source: Python, Node.js, Graphics card
Disk space: 1.0 GB–4.8 GB · estimate
Accounts or keys: None
Cost: Free
License: AGPL-3.0 (Copyleft)
Can replace: Midjourney, Firefly, Photoshop
Last update: Jul 27, 2024

Risk signals (5)

AttentionDependencies with known vulnerabilities
Pinned dependencies with known vulnerabilities (OSV): gradio: GHSA-26jh-r8g2-6fpr, GHSA-279j-x4gx-hfrh, GHSA-34rf-p3r3-58x2; protobuf: GHSA-7gcm-g887-7qv7, GHSA-8gq9-2x98-w8hf, GHSA-8qvm-5x2c-j2w7; transformers: GHSA-29pf-2h5f-8g72, GHSA-37mw-44qp-f5jm, GHSA-37q5-v5qm-c9v8; setuptools: GHSA-5rjg-fvgr-3xxf, GHSA-cx63-2mw6-8hw5, GHSA-h35f-9h28-mq5c; gitpython: GHSA-239g-whfq-7xj9, GHSA-23mf-xhv8-69c2, GHSA-284h-m62q-gf8w; pillow: GHSA-22hg-r758-rrw9, GHSA-24rc-8hj2-v98c, GHSA-254f-3936-288h; accelerate: GHSA-4j2p-28q2-5m79, PYSEC-2026-3804; diskcache: GHSA-w8v5-vhqr-4h9v, PYSEC-2026-2447; fastapi: PYSEC-2024-38; gradio: GHSA-26jh-r8g2-6fpr, GHSA-279j-x4gx-hfrh, GHSA-34rf-p3r3-58x2; protobuf: GHSA-7gcm-g887-7qv7, GHSA-8gq9-2x98-w8hf, GHSA-8qvm-5x2c-j2w7; pytorch_lightning: GHSA-4cv3-v7pv-rfhf, GHSA-75m9-98v2-hjpm, GHSA-98fp-7v67-4v3q; transformers: GHSA-29pf-2h5f-8g72, GHSA-37mw-44qp-f5jm, GHSA-37q5-v5qm-c9v8
automatic check
Attentioninactive development
Last commit 2024-07-27 (824 days before this check) and last release v1.10.1 on 2025-02-09, with 2510 open issues; the pinned dependencies are not being updated (see known_vulnerabilities).
AI review
Infoold python version required
README.md, "Automatic Installation on Windows": "Install Python 3.10.6 (Newer version of Python does not support torch), checking \"Add Python to PATH\"".
AI review
Infothird party extensions run code
README.md: "Custom scripts with many extensions from community" and "Running arbitrary python code from UI (must run with `--allow-code` to enable)". Extensions are code from other authors that runs on your PC.
AI review
Infoanime tagging for adult boorus
README.md: "DeepDanbooru integration, creates danbooru style tags for anime prompts". The models and tag sets people use with it come from outside this repository and can include adult material; the repository itself ships no models.
AI review

Evidence is shown in its original language.

What Radar could not check

Whether it still installs cleanly today, since development stopped in 2024 and the dependencies are pinned to old versions.
Real disk use after downloading torch and image models (the checks estimate 1 to 5 GB before extra models).
Speed and memory use on your graphics card; the README cites 4 GB cards and reports of 2 GB working.
What code the installer and the community extensions download at run time.
Which models you end up using, since the repository ships none and each has its own licence.