llama.cpp
#3 in Lokale LLM-RuntimesGeorgi Gerganov · siet March 10, 2023 · 3× · tolest 09. Okt. 2026
75
Momentum
llama.cpp is an open-source C/C++ library for LLM inference, originally developed by Georgi Gerganov and now maintained within the ggml-org project. It runs locally on CPU and GPU and forms the basis of many local LLM tools such as Ollama and LM Studio. The official website llama.app provides an installer and getting-started documentation.
Momentum-Verloop
11.07.09.10.
Features
| License | MIT |
| Platform | Windows, macOS, Linux (Docker, Homebrew) |
| Price | Free (open source) |
| Protocol Compatibility | OpenAI-compatible API (/v1/chat/completions, /v1/embeddings), Anthropic Messages-compatible, Ollama shim (/api/tags) |
| Release Date | March 10, 2023 |
| Supported Models/Providers | GGUF models (quantized), e.g. Qwen, Llama, Mistral, Gemma |