# llama.cpp: self-hosting guide & alternatives | SelfHostAI

Source: https://ai.andreiturcea.com/tool-llama.cpp.html
Last updated: 2026-10-15


**Short answer:** How to self-host llama.cpp: requirements, licence and alternatives.

Last updated: 15 October 2026 · No tax figures on this site, so no GOV.UK check applies; tool details come from each project's public GitHub repository · [Markdown version](https://ai.andreiturcea.com/tool-llama.cpp.html.md)

LLM inference in C/C++, runs on CPU.

| Category | LLM runtime |
|---|---|
| Licence | MIT |
| Repo | [ggml-org/llama.cpp](https://github.com/ggml-org/llama.cpp) |

## How do you self-host it?

See the official README for Docker or binary install. A 2 vCPU / 4 GB VPS runs most UIs; LLM runtimes need a GPU or must run small quantised models on CPU.

## What are the alternatives?

Ollama, LocalAI, vLLM