

llama.cpp
#3 in Local LLM RuntimesGgml · since 13. November 2026 (v1.0.0); Projektstart März 2023 · 2× · last seen Nov 17, 2026
39
Momentum
llama.cpp is a C/C++ inference engine that runs large language models (and vision-language models) locally on consumer hardware and in the cloud, mostly in GGUF format. The project is built on the ggml tensor library and developed by ggml-org (Hugging Face). With v1.0.0 (November 13, 2026), the public C API follows semantic versioning, and breaking changes require a major-version bump.
Momentum trend
24.08.22.11.
Features
| License | MIT |
| Platform | Windows, macOS, Linux; Docker; Homebrew, Nix, MacPorts |
| Price | Free; no subscriptions, accounts or usage fees |
| Protocol Compatibility | OpenAI-compatible API (llama-server, /v1); Anthropic-compatible endpoint; Ollama-compatible shim |
| Release Date | v1.0.0: November 13, 2026 |
| Supported Models/Providers | More than 150 model architecture variants (per release post); GGUF format |