Llama (Meta)
Also known as: Llama models · Llama 3
Llama is Meta’s family of open-weight LLMs. Because the weights are downloadable, Llama has become one of the most-used starting points in the local-model world.
- Run it on your own hardware via Ollama, llama.cpp, etc.
- A huge ecosystem of fine-tunes, quantizations, and tooling.
- Check the license — open weights ≠ open source, and terms vary by size.
Why it matters to you: if you run a local model, there's a good chance it's a Llama (or a derivative). "Llama" is to local models what "x86" is to CPUs — a default that everything else measures against.