# TextGen

TextGen is a desktop application for running local LLMs with text, vision and tool-calling support, offering a chat UI plus an OpenAI and Anthropic-compatible API entirely offline.

Portable builds for Linux, Windows and macOS bundle everything needed to run GGUF (llama.cpp) models with CUDA, Vulkan, ROCm or CPU-only backends, with no separate dependency install. It supports instruct and character-based chat modes with templated prompt formatting, image attachments for visual understanding, text, PDF and DOCX file attachments, message editing, and conversation branching, plus a free-form notebook tab outside of chat turns.

Additional backends, including ExLlamaV3, Transformers and TensorRT-LLM, can be switched without restarting, and tool-calling supports web search, page fetching, custom Python-file tools and MCP servers. A separate tab handles LoRA fine-tuning and diffusion-model image generation, and the app runs with zero telemetry or external resource calls.

Python, AGPL-3.0 licensed.

Website: https://github.com/oobabooga/textgen