# textgen review, pricing and alternatives

Source: https://listme.name/software/textgen
Website: https://github.com/oobabooga/textgen

Open-source desktop app for running local language models

textgen is an AGPL-3.0 desktop app for running local language models, with text, vision and tool-calling support. It offers an OpenAI- and Anthropic-compatible API, ships portable builds for Linux, Windows and macOS, and says it has no telemetry.

textgen is a desktop app for loading and running local language models. Its README lists text and vision models, tool-calling and web search, and it offers an API compatible with OpenAI and Anthropic clients. The portable builds cover CUDA, Vulkan, ROCm and CPU-only setups on Linux, Windows and macOS.

The repository is AGPL-3.0 licensed and had about 47,700 stars when checked. The README says the project is open source with no telemetry. Users who want a chat-first app may compare LM Studio or Jan, which are also in this ranking.

- Free and open source | Linux, Windows, macOS | open source (AGPL-3.0)
- Price: AGPL-3.0 desktop app is free; models are downloaded separately
- Checked: 2026-10-10

## Strengths

- AGPL-3.0 open source with no telemetry
- Portable builds for Linux, Windows and macOS
- OpenAI and Anthropic compatible API

## Limits

- AGPL-3.0 copyleft applies to modified versions run as a service
- Vision and tool features depend on the model used

## Rankings

- #7 of 10 in [best Local LLM runners](https://listme.name/best/local-llm-runners), score 82.5: AGPL-3.0 desktop app for local models with text, vision and tool-calling support, an OpenAI- and Anthropic-compatible API and portable builds for Linux, Windows and macOS. It suits people comfortable with local setup.

## Alternatives

- [Ollama](https://listme.name/software/ollama), score 92.0
- [llama.cpp](https://listme.name/software/llama-cpp), score 90.0
- [LM Studio](https://listme.name/software/lm-studio), score 89.0
- [GPT4All](https://listme.name/software/gpt4all), score 87.5
- [Jan](https://listme.name/software/jan), score 86.0
- [KoboldCpp](https://listme.name/software/koboldcpp), score 84.5
- [llamafile](https://listme.name/software/llamafile), score 81.0
- [MLX LM](https://listme.name/software/mlx-lm), score 80.0