What is LlamaGPT used for?

LlamaGPT is a self-hosted, offline chatbot that runs Llama 2 and Code Llama models on your own hardware, with an OpenAI-compatible API and no data leaving the device. It is a good fit for private chat assistant on a home server and local code assistant with Code Llama.

Is LlamaGPT open source?

Yes. LlamaGPT is open source under the MIT licence. Its source code is on GitHub at getumbrel/llama-gpt and is written mainly in TypeScript.

Yes. LlamaGPT is open source, so the software itself is free to use.

Can I self-host LlamaGPT?

Yes. LlamaGPT can be self-hosted on your own server or infrastructure; there is no official hosted version.

What is LlamaGPT an alternative to?

LlamaGPT is an open-source alternative to ChatGPT. Other open-source alternatives to ChatGPT include OpenClaw, Open WebUI and LobeHub.

Is LlamaGPT actively maintained?

The most recent commit to LlamaGPT was on 23 April 2024. The project has 11k stars on GitHub.

LlamaGPT: open-source alternative to ChatGPT

About LlamaGPT

LlamaGPT is a self-hosted chatbot that runs open Llama 2 language models locally, giving a ChatGPT-style chat interface without sending conversations to an outside service. It comes from the team behind Umbrel and is aimed at people who want a private assistant on a home server, a Mac or any Docker host.

It supports several models, including Nous Hermes Llama 2 in 7B, 13B and 70B sizes and Code Llama variants in 7B, 13B and 34B sizes, with download sizes and memory requirements listed in the README. Nvidia GPU support was added, and an OpenAI-compatible API lets existing client code point at the local instance. Running custom models is listed on the roadmap rather than available today.

Installation is a one-click app on umbrelOS, a Docker-based setup on an M1 or M2 Mac, or a Docker or Kubernetes deployment elsewhere. It is MIT licensed and runs entirely on your hardware, with the chat interface opened in a browser on the local machine.

Key features

Offline ChatGPT-style chat interface
Runs Llama 2 and Code Llama models
OpenAI-compatible API endpoint
Nvidia GPU support
One-click install on umbrelOS
Docker and Kubernetes deployment options

Good fit for

→Private chat assistant on a home server
→Local code assistant with Code Llama

Built with: TypeScript
Tags: llm; chatbot; llama; self-hosted; offline; openai-compatible; docker; privacy

LlamaGPT: questions and answers

What is LlamaGPT used for?: LlamaGPT is a self-hosted, offline chatbot that runs Llama 2 and Code Llama models on your own hardware, with an OpenAI-compatible API and no data leaving the device. It is a good fit for private chat assistant on a home server and local code assistant with Code Llama.
Is LlamaGPT open source?: Yes. LlamaGPT is open source under the MIT licence. Its source code is on GitHub at getumbrel/llama-gpt and is written mainly in TypeScript.
Is LlamaGPT free?: Yes. LlamaGPT is open source, so the software itself is free to use.
Can I self-host LlamaGPT?: Yes. LlamaGPT can be self-hosted on your own server or infrastructure; there is no official hosted version.
What is LlamaGPT an alternative to?: LlamaGPT is an open-source alternative to ChatGPT. Other open-source alternatives to ChatGPT include OpenClaw, Open WebUI and LobeHub.
Is LlamaGPT actively maintained?: The most recent commit to LlamaGPT was on 23 April 2024. The project has 11k stars on GitHub.