Text Generation WebUI

Text Generation WebUI is a desktop application that runs local large language models. It supports text-based chat and generation, vision capabilities, and tool-calling, with an API-compatible interface for local use. The software is designed for users who want private, offline operation and direct control over their models and workflows.

It provides portable builds that run on major operating systems, offering CUDA, Vulkan, ROCm, and CPU-only options. The UI aims to be easy to start with, and the project emphasizes privacy and offline operation, enabling users to work without external telemetry or remote updates.

Users can download GGUF model files, place them in a user_data/models folder, and let the UI detect them automatically. The software can be run in a browser-like window or via a local server, with multiple backends and model options available to switch without restarting.

Full installation options are provided for additional backends, training, and extensions, along with Docker setups for different GPUs. The project covers installation, startup scripts, and how to update or extend the system, all aimed at enabling local, private LLM use.

Use cases

• Local development with private LLMs
• Multimodal conversations