If you need AI product photos, room mockups, or social images but do not want a per-image subscription, AUTOMATIC1111’s Stable Diffusion WebUI is a proven free option. It is older and less polished than newer tools, but it runs in a browser, works on modest hardware, and has a huge community. This tutorial covers installing it and generating your first images.
What It Is
AUTOMATIC1111’s Stable Diffusion WebUI is a free, open-source web interface for Stable Diffusion, built with the Gradio library. You install it on your own computer, open it in a browser, type a prompt, and generate images locally. The project is published under the GNU Affero General Public License v3.0 and, as checked on October 17, 2026, has 165,071 stars on GitHub (github.com/AUTOMATIC1111/stable-diffusion-webui). Its official description is simply: “Stable Diffusion web UI.” One thing to know before you invest time: the repository’s last code push was on March 2, 2026, so development has slowed. Check the repository for the latest status before building a long-term workflow around it.
Who It’s For / SMB Use Case
This tool fits businesses that want a simple, free way to generate images on hardware they already own. Practical uses include:
- Product and menu visuals: Generate background scenes, mood boards, and draft product shots for social posts and ads.
- Mockups and variations: Test many color, background, and style variations of one idea before paying a designer.
- Low-budget content: Produce images on a 4GB graphics card, which the README says is supported, with reports of 2GB cards working.
- Private drafts: Keep unreleased product images on your own machine instead of uploading them to a hosted service.
This tool is not ideal if you want the newest models, a guided setup, or a maintained commercial product. Newer interfaces such as ComfyUI are more actively developed. Choose AUTOMATIC1111 when you want the simplest free local image tool with a familiar prompt box and lots of community guides.
Installation
The README says you must install Python and Git first, and points to separate wiki pages for NVIDIA, AMD, Intel, and Ascend hardware. Pick the path below that matches your system. Commands and steps come from the official README.
Windows 10/11 with NVIDIA GPUs (release package)
- Download
sd.webui.zipfrom the v1.0.0-pre release and extract its contents. - Run
update.bat. - Run
run.bat.
Automatic installation on Windows
- Install Python 3.10.6, checking “Add Python to PATH”. The README warns that newer versions of Python do not support torch, so use the exact version listed.
- Install Git.
- Download the repository, for example by running:
git clone https://github.com/AUTOMATIC1111/stable-diffusion-webui.git
- Run
webui-user.batfrom Windows Explorer as a normal, non-administrator user.
Automatic installation on Linux
- Install the dependencies. On Debian-based systems:
sudo apt install wget git python3 python3-venv libgl1 libglib2.0-0
The README also gives commands for Red Hat, openSUSE, and Arch-based systems, and notes that very new systems may need Python 3.11 or 3.10. Check the README for the command matching your distribution.
- From the directory where you want the WebUI installed, run:
wget -q https://raw.githubusercontent.com/AUTOMATIC1111/stable-diffusion-webui/master/webui.sh
Or clone the repository directly:
git clone https://github.com/AUTOMATIC1111/stable-diffusion-webui
- Run
webui.sh. - Check
webui-user.shfor options.
Apple Silicon
For Apple Silicon Macs, follow the instructions on the project’s wiki page linked from the README: Installation on Apple Silicon.
How to Use AUTOMATIC1111
Step 1: Open the WebUI and pick a checkpoint
When the server starts, the terminal prints the local address to open in your browser. The WebUI opens on a tabbed page. Choose a checkpoint model from the dropdown at the top left. You need to download a checkpoint model separately and place it in the models folder; check the official docs or wiki for the current model folder layout and a compatible model source.
Step 2: Generate your first image with txt2img
Go to the txt2img tab. Type a plain description in the prompt box, for example: “cozy boutique hotel lobby, plants, warm light, photorealistic”. Type things you want to avoid in the negative prompt box, such as “blurry, distorted, low quality”. Set the image size, then click Generate. The README notes that generated images’ dimensions must be a multiple of 8, and that generation parameters are saved with each image.
Step 3: Control emphasis with attention syntax
Use the attention syntax from the README to tell the model which words matter most. Wrapping a word in double parentheses increases attention:
a man in a ((tuxedo))
You can also set an explicit weight:
a man in a (tuxedo:1.21)
Select text and press Ctrl+Up or Ctrl+Down (or Command+Up / Command+Down on macOS) to adjust attention to the selected text automatically.
Step 4: Rework an existing photo with img2img
Go to the img2img tab, upload a photo, and describe the change you want. Use a low denoising strength for small touch-ups and a higher value for bigger changes. Batch Processing in img2img can process a group of files at once, which is useful when you need consistent edits across a set of product photos.
Step 5: Upscale, fix faces, and reuse settings
Use the Extras tab for upscaling with tools listed in the README, including RealESRGAN and ESRGAN, and for face restoration with GFPGAN or CodeFormer. Use Highres Fix for higher-resolution output in one click. To reproduce an image, drag it onto the PNG Info tab to restore the generation parameters and copy them back into the UI.
Pricing / Limitations
The WebUI itself is free and open source under the AGPL-3.0 license. Your costs are hardware, electricity, disk space, and the time to set it up. A 4GB graphics card is supported, per the README, but generation will be slower than on a stronger GPU. The biggest limitations are the slowed development pace, the fiddly setup (especially the exact Python version on Windows), and the fact that checkpoint models each carry their own licenses, so check a model’s license before using images commercially. Do not expose the WebUI directly to the public internet; check the official docs and wiki for safe hosting guidance first. If the README does not state something clearly, check the official wiki at github.com/AUTOMATIC1111/stable-diffusion-webui/wiki.
Bottom Line for SMB
AUTOMATIC1111 is a solid free starting point for local AI image generation, especially on modest hardware. Install it, generate a batch of drafts, and keep the winners. If you need newer models or a more active project, look at ComfyUI next. Either way, strong prompts matter more than the tool itself, so see our AI prompt templates for small business and our guide to per-seat vs. usage-based AI pricing before you commit budget to any image tool.
Frequently Asked Questions
Is AUTOMATIC1111 still maintained?
Development has slowed. As checked on October 17, 2026, the last code push to the repository was March 2, 2026. Check the repository for the latest status before relying on it for a long-term workflow.
How much VRAM do I need?
The README says 4GB video cards are supported, with reports of 2GB cards working. More VRAM means larger images and faster generation.
Can I use generated images in my advertising?
The WebUI is free under the AGPL-3.0 license, but each checkpoint model has its own license. Check the license of the model you used before putting images in paid advertising.
