llama.cpp
A C++ inference engine that runs large language models locally from quantized GGUF files, on CPU, Apple Silicon or a consumer GPU.
- Repository
- ggml-org/llama.cpp
- Category
- local-llm
- Last verified
- October 14, 2026
GitHub
- License
- MIT
- Language
- C++
- Last push
- Oct 15, 2026
- Latest release
- v0.7.0 (Oct 15, 2026)
Stats as of Oct 15, 2026
Why it matters
The cheapest way to run a serious model is a quantized one on hardware you already own — and this is the engine that makes that possible on a mid-range machine.
- llm
- inference
- gguf