Skip to content
buildbay.
SD

Stable Diffusion web UI

AUTOMATIC1111 · Images

GitHub stars

165,173
—Collecting 30d data

Stats updated

165,139 stars on Sep 27 to 165,173 stars on Oct 1, up 34.5 measured star snapshots

What is Stable Diffusion web UI?

Stable Diffusion web UI assembles a wide range of local generation, image-revision, prompt, model, and training controls behind a Gradio interface. Its tabs and settings support direct experimentation, batch work, metadata-based reproduction, extension scripts, and API access without requiring users to build a custom interface around the underlying models.

Stable Diffusion WebUI is an open-source web interface for Stable Diffusion maintained in the AUTOMATIC1111 repository. Its documentation includes installation paths for NVIDIA, AMD, Intel, Ascend, and Apple Silicon hardware.

Core capabilities

  • Run text-to-image, image-to-image, inpainting, outpainting, and upscaling
  • Control prompts, seeds, samplers, checkpoints, VAEs, LoRAs, and embeddings
  • Recover generation settings from media and extend workflows through scripts or API

Generation controls in one UI

The interface places text-to-image and image-to-image beside inpainting, outpainting, high-resolution fixes, prompt weighting, styles, variations, seed controls, checkpoint switching, VAE selection, and multiple upscalers. Users can interrupt work, preview progress, and batch-process inputs from the same local service.

  • txt2img
  • img2img
  • inpainting
  • upscaling

Reproducible and extensible

Generation parameters can be stored in PNG or JPEG metadata and loaded back into the prompt controls, providing a path from an output to its settings. Custom scripts, community extensions, learned embeddings, LoRAs, training tools, and an API broaden the base interface for specialized local workflows.

  • metadata
  • extensions
  • LoRA
  • API

Where it fits

Use cases

  1. 01

    Explore prompt and model settings

    Compare samplers, seeds, checkpoint or VAE choices, prompt weights, styles, and other generation parameters from one interface while previewing progress and preserving the settings used for an output.

  2. 02

    Repair or enlarge an image

    Use image-to-image, masks, inpainting, outpainting, face restoration, and supported upscalers to revise selected content or prepare a higher-resolution result from an existing source image.

  3. 03

    Assemble a local extension stack

    Install custom scripts and extensions, organize model files and learned embeddings, enable the API, and tailor defaults or interface controls around the generation workflow used on a particular workstation.