No API key. No subscription. No data leaving your machine. Pick your use case, hardware, and preferred runtime — get the right setup with exact commands.
01 — What do you want to do?
02 — How much RAM?
03 — GPU type?
04 — Preferred runtime?
Meta's Llama 3.2 8B. Excellent instruction following — a solid all-rounder for everyday chat and writing.
ollama run llama3.2:8b
Complete Stack
Pull and run any open-source model with one command. Manages weights, versioning, and exposes a local REST API compatible with the OpenAI SDK.
curl -fsSL https://ollama.com/install.sh | sh
API server at http://localhost:11434/v1
Self-hosted ChatGPT-style UI for Ollama. Supports RAG, vision, voice, multi-user, and image generation — runs entirely in your browser.
docker run -d -p 3000:8080 \ --add-host=host.docker.internal:host-gateway \ -v open-webui:/app/backend/data \ --name open-webui \ ghcr.io/open-webui/open-webui:main
Open at http://localhost:3000
New to Ollama?
Read our complete Ollama setup guide — install steps, model comparisons, and tips for every OS.