What is Ollama?

Ollama is a free, open-source tool that lets you download and run AI models on your own computer with a single command. Type ollama run plus a model name in your terminal, and it fetches the model and starts a chat — no cloud account, no API key, no per-token bill.

It is the most popular on-ramp for running an LLM locally. Under the hood it wraps llama.cpp, the engine that actually crunches the numbers, and handles the fiddly parts for you: picking a quantized (shrunk-down) version of the model, using your GPU if you have one, and managing downloads.

Its model library covers most of the big open-source LLM names — Llama, Qwen, Gemma, DeepSeek, gpt-oss — in sizes from tiny models that run on an ordinary laptop to heavyweights that need a serious GPU.

Two things make it especially developer-friendly. It runs a local server that mimics OpenAI's API, so tools built for ChatGPT can point at your own machine instead. And because it is command-line first, it slots neatly into scripts and automations.

The catch: your hardware sets the ceiling. A small model running on a laptop will not match the frontier models you rent through an API, and bigger models need lots of RAM. But for privacy, offline use, or plain tinkering, Ollama is the easiest starting point — and if you would rather click than type commands, LM Studio covers the same ground with a friendlier interface.