Ollama

Ollama is an open source tool that lets you run large language models on your own hardware. You install it on macOS, Windows, or Linux, then pull models by name and run them with simple commands. It handles model weights, configurations, and GPU acceleration so you can focus on using the models rather than configuring them.

The tool supports a library of over 100 models including Llama, Mistral, Qwen, and others. You can run multiple models at the same time, and it manages memory allocation between them. An OpenAI-compatible API means you can swap local models into existing applications with minimal code changes.

Ollama added optional cloud features in 2025, letting you access larger models on datacenter hardware when your local machine is not powerful enough. Local use remains free. The Pro tier costs $20 per month and the Max tier costs $100 per month for higher cloud usage limits. Some users have raised concerns about the cloud direction, preferring the tool stay focused on local-only functionality.