bash — ~/projects
$curl -fsSL https://ollama.com/install.sh | sh
>> Installing Ollama...
>> Installation complete ✓
 
$ollama run llama3
>> pulling manifest...
>> pulling 4.7 GB [████████████████] 100%
>> success
 
Send a message (type /bye to exit)
 
Explain quantum computing in simple terms
Quantum computing is like having a magical coin that can be heads AND tails at the same time...
$ open source · local first

Get up and running with
large language models.

Run Llama 3, Mistral, Gemma, and other models locally on your machine. No API keys, no cloud dependency — just you and your GPU.

# Why Ollama?

Simple. Local. Open source. The way AI should be.

🖥️

$ ollama run

One command to download and run any supported model. No configuration files, no complex setup.

$ ollama run mistral
🔒

100% Local

Your data never leaves your machine. Complete privacy and control over model inference.

# no API calls, ever

Built-in Modelfile

Create custom models from a simple config file. Import GGUF, set parameters, define system prompts.

$ ollama create mymodel -f Modelfile
🔌

REST API

Compatible with OpenAI's API format. Drop-in replacement for existing integrations and tools.

$ ollama serve → localhost:11434

# Get Started

No pricing tiers. No subscriptions. Just free & open source software.

Self-hosted

DIY

run it anywhere

  • Docker support
  • Kubernetes ready
  • Custom model serving
  • Team deployment
Deploy Guide

Ecosystem

community powered

  • Open WebUI
  • LobeChat integration
  • LangChain / LlamaIndex
  • VS Code extension
Explore Tools