---
id: "b636"
language: "en"
library: "tools"
official_url: "https://github.com/ggml-org/llama.cpp"
profile_updated_on: "2026-10-22"
---

# llama.cpp

Run local models in C/C++, with command-line tools, quantization utilities and a model server with a web interface.

- Company / organization: ggml-org
- Category: Model development
- Official website: [https://github.com/ggml-org/llama.cpp](https://github.com/ggml-org/llama.cpp)
- Profile updated: 2026-10-22


## Overview

llama.cpp is a model-inference project maintained by ggml-org, not a single Llama model. With compatible model files, users run a chat CLI or start a server with a web interface and compatible APIs. The project supports GGUF workflows, quantization and multiple CPU/GPU backends for local or self-managed inference. Check model files, licenses and hardware support separately; the project name does not imply every model is compatible.

## Official sources and references

- [Official website](https://github.com/ggml-org/llama.cpp)
- [Documentation](https://github.com/ggml-org/llama.cpp/tree/master/tools/server)

## Reading & subscriptions

- [Interactive profile](../../../products/llama-cpp--b636.html)
- [Read](../../../read/en/products/llama-cpp--b636.html)
- [JSON](../../../read/en/products/llama-cpp--b636.json)

Descriptions summarize public information; consult official websites for features, pricing and regional availability.