Open Interpreter
Open source terminal coding agent built for low-cost and open-weight models.
- SI Score
- 80/100
- Pricing
- Not confirmed
- Free plan
- Not confirmed
- Sign-up required
- Not confirmed
Ask SI about Open Interpreter
Answers use saved listing facts and recorded sources.Homepage screenshot
What it does
Open Interpreter is an open source coding agent that runs in the terminal. After installing the interpreter command, you can open a project, pick a provider and model, and have the agent read files, edit code, and run commands. It ships model-specific harnesses for Kimi, Qwen, DeepSeek, and other open models, and it is provider agnostic, with built-in providers including OpenAI, Amazon Bedrock, Ollama, and LM Studio. Sandbox mode and approval policy control what runs automatically and what needs approval. The project is licensed under Apache 2.0 and is a fork of Codex. The site also links to a desktop app for Word, Excel, PDFs, browsers, and local files.
Install with the shell command shown on the site: curl -fsSL https://www.openinterpreter.com/install | sh. Documentation covers interactive mode, CLI reference, sessions, slash commands, sandbox and approvals, permissions, MCP server, AGENTS.md, skills, and non-interactive mode. Sandboxing works on macOS, Linux, and Windows, and network access starts disabled in permission profiles unless enabled. Local models stay on your machine; with your own API key requests go directly to the provider. Hosted models route through the project's servers and are logged for up to 30 days. An OpenAI SDK, app server, and agent client protocol are also documented.