What Ollama does

Ollama provides a way to run supported AI models on your own computer and use them from an application or a developer workflow. You install the software, select a model that fits your hardware, and interact with it through the app, terminal or API. That makes it a useful starting point for local AI experimentation and integrations that need a model endpoint. The Ollama software is MIT-licensed and local model execution does not require a subscription. The company also offers paid cloud model usage and plans. These are different execution routes: local performance depends on your hardware, while cloud requests are processed by a hosted service.

Best for

  • Trying supported AI models locally
  • Developers building a local model integration
  • Connecting model backends to development tools

Key features

  • Local model execution on major desktop systems
  • App, terminal and API workflows
  • Optional cloud execution for hosted models

Things to consider

  • Local model size and speed depend on available memory and hardware.
  • Individual model licenses remain separate from Ollama's software license.
  • Cloud models use a hosted service and have their own charges.

Getting started

  1. Install Ollama for your operating system.
  2. Select a local model suitable for the computer.
  3. Try a small task, then connect the API or a supported application if needed.