llama.cpp
Georgi Gerganov's legendary C/C++ port of LLaMA inference. Powers most local LLM tools on the planet — CPU/GPU, GGUF, quantization, and full offline operation.
Tags
Open Source · Free · Local · Self-Hostable · Offline · 2026 · llama.cpp · llama-cpp · open source ai · free ai tool · local ai · local model