Local model tools
Ollama on GitHub: what it does and how to use it with AI agents
Run open models like Qwen, DeepSeek, gpt-oss and Gemma on your own machine.
Stars, license and last push from the GitHub API, checked 2026-10-26
| Repo | ollama/ollama |
|---|---|
| Stars | 184,800+ |
| License | MIT |
| Last push | 2026-10-26 |
| Language | Go |
What Ollama does
Ollama downloads and runs open models locally with one command and serves them on a local API. Agents and editors that accept a custom endpoint can use it instead of a paid API.
Who it is for
Anyone who wants private or offline AI, or cheap local models for simple agent steps.
How to use Ollama with Claude Code, Codex or Cursor
Install from ollama.com or the GitHub releases.Run a model: ollama run qwen3 (or any model in the library).Point a tool at http://localhost:11434, the local API.Use local models for small tasks and keep hard coding work on a frontier model.
Tip: Check your RAM against the model size before you pull. Large models will not fit on most laptops.
The loadout for this repo
See the full tool stack, prompts and costs in Best setup to run AI locally.
Related repos
- llama.cpp: The C/C++ engine for running LLMs locally on almost any hardware.
- MCP Servers: The official reference servers for the Model Context Protocol.
Ollama FAQ
Is Ollama free?
Yes. It is open source under MIT and runs on your own hardware.
Ollama vs llama.cpp?
Ollama is the easier app built for one-command use. llama.cpp is the lower-level engine with more tuning options.
Can coding agents use Ollama?
Tools that accept a custom OpenAI-compatible or Ollama endpoint can. Check each tool's docs.
Star counts change daily. Always read a repo's license and code before running it. All repos