openinterpreter
ActiveDescription
Open Interpreter lets LLMs execute code and control a computer locally, turning natural language into shell, Python, and browser actions on the user machine.
Key Features
- Coding agent optimized for low-cost open-source models like DeepSeek, Kimi, and Qwen
- Multiple harness emulations including native, claude-code, swe-agent (9 harnesses)
- Native sandboxed execution on macOS, Linux, and Windows
- Built-in QA skill to drive agent-browser and trycua for web and GUI testing
- Local-first design with config and sessions stored under `~/.openinterpreter`
- Full support for MCP, Skills, Hooks, Permissions, and AGENTS.md
Use Cases
💡 Run a zero-cost coding agent locally with open-source LLMs
💡 Switch between harnesses to get the most out of different model capabilities
💡 Let the agent operate real desktop and browser apps for end-to-end testing
💡 Use as an open-source alternative to Claude Code
Categories
Quick Start
# One-line install for macOS / Linux
curl -fsSL https://openinterpreter.com/install | sh
# Launch interactive terminal
interpreter
# Switch model and harness in the TUI
> /model claude-sonnet-4
> /harness claude-code
> Analyze this CSV and plot a chart