Open Interpreter is now a Rust-based coding agent forked from OpenAI's Codex CLI, designed for open and low-cost models such as Kimi and GLM. The original Python project, which let an LLM run code on your computer, is maintained separately as a community fork.
It runs commands in native sandboxing on macOS, Linux and Windows, and works with any OpenAI-compatible provider through Chat Completions. Switch models with /model and harnesses with /harness. It reuses AGENTS.md and .agents/skills, supports MCP, hooks and permissions, and runs as an Agent Client Protocol agent for editors through interpreter acp.
Install on macOS or Linux with curl -fsSL https://www.openinterpreter.com/install | sh, or use the PowerShell script on Windows, then run interpreter or i. It is Apache-2.0 licensed.