❓ How Does it Work?

Open Interpreter equips a function-calling language model with an exec() function, which accepts a language (like "Python" or "JavaScript") and code to run.

We then stream the model's messages, code, and your system's outputs to the terminal as Markdown.