Ollama ↗
Get a language model running on your own computer, and start building locally.
Self-hosted
FREEAI GO · INDEPENDENT DIRECTORY
llama.cpp: Free Features, Limits & Alternatives
Run language models efficiently across a wide range of local hardware.
What is actually free?
The inference engine is MIT-licensed. You need compatible model files, enough memory, and your own compute; model licenses vary.
Official documentation checked: 2026-10-11. Not a full product test. Terms and availability can change.Our editorial assessment
We rate documented usefulness, ease of starting, and the scope of free access, each out of five. The displayed score is their equal-weight average. These are editorial opinions, not user reviews or hands-on benchmarks.
Before you start
This is a self-hosted tool. Check operating-system support, model licensing, hardware requirements, and any external API costs before installation.