How do you run a local LLM?

The fastest way to run a local LLM in 2026, per Ollama's official README:

1. Install Ollama for your OS.
2. Open a terminal and run ollama pull llama3.2 (see the Llama 3.2 model page for tags).
3. Run ollama run llama3.2 and start chatting.

For a GUI, try LM Studio or Jan. For IDE integration, point Cursor or Continue at your local OpenAI-compatible endpoint — Ollama serves one on localhost:11434 per Ollama's API docs.

Check RAM against model size listings: 8 GB can run small 3B models; 16 GB is comfortable for 7B–8B; 32 GB+ opens 14B and some 32B quantized models.