Repolode
Submit
oobabooga

oobabooga/textgen

A free, open-source desktop app for running large language models locally, with a chat UI and an OpenAI/Anthropic-compatible API.

Open-source alternative to:LM Studio

View on GitHub
Stars
47,739
Forks
5,987
Open issues
865
Last commit
2 months ago

What is it?

TextGen is a desktop application for running large language models on your own machine. It bundles a chat interface with an API, so the same local model can be used from a browser-style UI or from code. Portable builds are provided for Linux, Windows, and macOS, with CUDA, Vulkan, ROCm, and CPU-only options, and it runs GGUF models through llama.cpp. Additional backends such as ExLlamaV3, Transformers, and TensorRT-LLM are available through the full installation.

Beyond plain chat, it supports instruct and character-based chat modes, image attachments for vision models, and text and PDF file uploads. Models can call tools such as web search, page fetching, and custom Python functions, and MCP servers are supported. The project also includes LoRA fine-tuning, a diffusers-based image generation tab, and extensions for TTS, voice input, and translation. The app states that it works fully offline and sends no telemetry.

The API exposes Chat, Completions, and Messages endpoints that follow the OpenAI and Anthropic formats, so existing client code can point at a local server instead.

Key features

  • 01Chat, instruct, and character modes with Jinja2 prompt templates
  • 02Vision input, plus PDF, .docx, and text file attachments
  • 03Multiple backends including llama.cpp, Transformers, and ExLlamaV3
  • 04OpenAI/Anthropic-compatible API with tool-calling support
  • 05Tool calling with custom Python functions and MCP servers
  • 06LoRA fine-tuning and a diffusers-based image generation tab

Use cases

  • Running a GGUF chat model privately on a personal computer without sending data to a cloud service
  • Pointing existing OpenAI or Anthropic client code at a local model during development
  • Chatting with documents or images using a local vision-capable model
  • Fine-tuning a LoRA adapter on your own chat or text dataset

Quick start

  1. 1

    Download the portable build for your OS from the releases page

  2. 2

    Unzip the download and double-click textgen to open the app window

  3. 3

    Place a GGUF model file in the user_data/models folder and the UI will detect it

Things to know

  • Licensed under AGPL-3.0, a copyleft license that requires derivative network services to share source code.
  • Additional backends such as ExLlamaV3 and Transformers, plus training and image generation, need the full install, which takes about 10GB of disk space and downloads PyTorch.
  • The portable build supports GGUF (llama.cpp) models only.

Topics

#local-llm#llm#llama-cpp#gguf#chat-ui#openai-compatible-api#lora-fine-tuning#image-generation

FAQ

What is textgen?
A free, open-source desktop app for running large language models locally, with a chat UI and an OpenAI/Anthropic-compatible API. TextGen is a desktop application for running large language models on your own machine. It bundles a chat interface with an API, so the same local model can be used from a browser-style UI or from code. Portable builds are provided for Linux, Windows, and macOS, with CUDA, Vulkan, ROCm, and CPU-only options, and it runs GGUF models through llama.cpp. Additional backends such as ExLlamaV3, Transformers, and TensorRT-LLM are available through the full installation.
Is textgen free to use? What license does it have?
textgen is open source under the AGPL-3.0 license. Check the license terms before using it in commercial products.
Is textgen still maintained?
Updated within 6 months. The last commit to textgen was on Aug 17, 2026.
What can I use textgen for?
Running a GGUF chat model privately on a personal computer without sending data to a cloud service Pointing existing OpenAI or Anthropic client code at a local model during development Chatting with documents or images using a local vision-capable model Fine-tuning a LoRA adapter on your own chat or text dataset
How do I install textgen?
Download the portable build for your OS from the releases page Unzip the download and double-click textgen to open the app window Place a GGUF model file in the user_data/models folder and the UI will detect it
Where can I get textgen?
The source code and installation instructions are on GitHub: https://github.com/oobabooga/textgen

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects