text-generation-webui

text-generation-webui, also called TextGen, runs language models on your own hardware for chat, writing and application integrations. You can use it as a desktop app or a self-hosted browser interface.

• Chat and writing: Talk to instruction-following models or custom characters, edit messages, branch conversations and generate free-form text in a notebook.
• Images and documents: Use vision models to discuss images and attach PDF, DOCX or text files to conversations.
• Local model choice: Run GGUF models through llama.cpp, or use backends including ik_llama.cpp, Transformers, ExLlamaV3 and TensorRT-LLM; switch models and backends without restarting.
• API and tools: Connect other applications through OpenAI- and Anthropic-compatible APIs, with tool calling for custom functions and MCP servers.
• Training and media: The full installation supports LoRA fine-tuning and image generation with diffusers models such as Z-Image-Turbo, plus extensions for speech, voice input and translation.

TextGen is open source under AGPL-3.0. Portable desktop builds support Windows, macOS and Linux with CPU-only, NVIDIA and AMD GPU options. The full installation also supports Apple Silicon, and Docker deployment is available. Portable builds use GGUF models; Transformers and EXL3 formats require the full installation.

Local generation works offline without telemetry. Web search and page fetching require internet access.