Open Source
Developer Tools Local LLMsCLILlamaPrivacyOffline
Ollama
Get up and running with LLMs like Llama 4, Gemma, and Phi on your Mac, Linux, or Windows machine — no cloud required.
Install
$curl -fsSL https://ollama.com/install.sh | sh About
Ollama bundles model weights, configuration, and runtime into a single binary. One CLI command pulls a model and starts a local server with an OpenAI-compatible API, so existing LangChain, LiteLLM, and OpenAI SDK code works unchanged.
The project supports Apple Silicon GPU acceleration via Metal and NVIDIA/AMD GPUs via CUDA/ROCm. It manages model files in ~/.ollama and exposes a REST API on port 11434.
Getting Started
# macOS / Linux
curl -fsSL https://ollama.com/install.sh | sh
# Pull and run a model
ollama run llama4
# Or via the REST API
curl http://localhost:11434/api/generate -d '{"model":"llama4","prompt":"Hello"}'Features
Run Llama 4, Gemma 3, Phi-4, Qwen2.5, Mistral, and 100+ models locally
OpenAI-compatible REST API on localhost:11434
Modelfile format for customising model behaviour and system prompts
GPU acceleration on Apple Silicon, NVIDIA, and AMD
Multi-modal support (vision models)
Automatic model management — pull, list, remove via CLI
Docker image available
100+ models: LLaMA, Mistral, Gemma, Qwen
One-command model pull & run
OpenAI-compatible REST API
GPU & CPU acceleration
Multi-model management
Cross-platform: Mac, Linux, Windows
References & Links
Project Info
Language Go
License MIT
Stars 132.0k ★
Forks 11.2k
Category Developer Tools
Topics
Local LLMsCLILlamaPrivacyOffline
Ready to try Ollama?
Free and open source. Install in one command, works with Claude Code, Cursor, Windsurf, and more.
$curl -fsSL https://ollama.com/install.sh | sh