Open Interpreter is a long-running open-source coding agent project with more than 68,000 stars on GitHub that lets language models run code on your computer to complete tasks.
One of the pioneering open-source agent projects, Open Interpreter gives an LLM a real execution environment: the model writes Python or shell code, runs it locally, sees the output and iterates until the task is done — summarizing files, converting formats, cleaning data, controlling applications and automating workflows in natural language. The current generation of the project is built in Rust and Python and positions itself as a coding agent for open models, with first-class support for models such as Kimi and GLM as well as DeepSeek and Qwen through its provider integrations. It speaks the Agent Communication Protocol (ACP), so it can plug into broader agent ecosystems. The project has appeared on GitHub Trending many times since 2023 and was still actively developed in October 2026 under the Apache License 2.0.
Open Interpreter helped define the category of computer-using agents years before it became an industry race, and its community produced a long line of demos — controlling apps, wrangling spreadsheets, automating research — that showed what happens when a model can act rather than only answer. Its newest generation leans into open-weight models, reflecting how capable they have become at agentic coding.
Open Interpreter is free and open source under the Apache License 2.0. You supply your own model access — a local model costs nothing, while hosted providers bill their own API usage.
Developers and power users who want a ChatGPT-style agent that can actually do things on their own computer, especially those who prefer open-weight models over closed hosted assistants.