# llama.cpp — AI Compare Ninja

> Open-source software that runs language models on your own computer, as a small app for Mac and Windows or from the command line.

## Overview
- **Category:** Local AI (Runtimes & servers)
- **Status:** Active · checked
- **Paid from:** Not verified for this market
- **Free access:** Free plan
- **Website:** https://llama.app/
- **Canonical URL:** https://www.aicompare.ninja/en/tool/llama-cpp

## What it does
- Free, open source and usable offline per the vendor; app for Mac and Windows 11, command line for Linux
- Per the project, a CPU is enough; graphics cards are used via Metal, CUDA and Vulkan, among others
- Local API that the vendor says is compatible with the OpenAI API

## In brief
llama.cpp is software for running language models, developed at ggml-org. The company ggml.ai supports its development and, per its website, has been part of Hugging Face since 2026. On llama.app there is a small app called Llama: it suggests models that fit your computer and opens a chat in the browser.

The software is free. Per the project page, it also runs on a CPU alone and can split models that are larger than the graphics memory between CPU and graphics card. Other programs can use it through the local API.


## Limits
Runs on your own computer; you only pay for hardware and power

## Pricing tiers
| Tier | Price | Period | What you get |
| --- | --- | --- | --- |
| Free | free | free | Runs on your own computer; you only pay for hardware and power |

## Sources
All details regarding plans, limits, and features are based on official vendor documentation and the primary sources verified below.

- https://llama.app/
- https://ggml.ai/
- https://github.com/ggml-org/llama.cpp

---
*Source: AI Compare Ninja (2026) - https://www.aicompare.ninja/en/tool/llama-cpp*