Ollama is an app that lets you quickly dive into playing with 50+ open source models right on your local machine, such as Llama 2 from Meta. It’s CLI-based, but thanks to the community, there are plenty of frontends available for an easier way to interact with the models. You can even customize a model to your specific needs pretty easily by adding a system prompt.

Open source models might not be as good as GPT-4 but they’re still pretty impressive. And, for those worried about data privacy, these models run locally, so your conversations stay on your computer.

Ollama also includes a REST API, making it a practical local alternative to the OpenAI API for coding projects.