OpenInterpreter

Open Interpreter CLI is an open source AI coding agent and terminal agent for macOS, Linux, and Windows.It runs in the terminal or as a desktop app, reads and edits project files, runs commands, and executes tests inside an OS-level sandbox with configurable permission profiles that require explicit approvals.

Model-specific harnesses support open-weight and hosted providers such as OpenAI, Anthropic, Amazon Bedrock, Ollama, LM Studio, Kimi, Qwen, and Deepseek, and custom providers compatible with OpenAI- or Anthropic-style APIs can be added.

Agent workflows use agents.md, reusable skills, and MCP servers to define project behavior, integrations with issue trackers, private docs, and databases.The CLI includes interactive commands (interpreter, resume, fork, side), project initialization, and developer docs for connecting models and building harnesses.

Open Interpreter is released under the Apache 2.0 license and supports local-only model execution for on-device inference while routing hosted-model requests through provider APIs.Use cases include code editing, test automation, debugging, CI assistance, and sandboxed team workflows with auditable approvals.

Key Features

• Cross-platform terminal and desktop agent for macOS, Linux, and Windows
• Reads and edits project files, runs commands, and executes tests inside an OS-level sandbox with configurable permission profiles and explicit approvals
• Model-specific harnesses supporting open-weight and hosted providers (OpenAI, Anthropic, Bedrock, Ollama, LM Studio, Kimi, Qwen, Deepseek) and custom providers compatible with OpenAI-/Anthropic-style APIs
• Agent workflows defined via agents.md, reusable skills, and MCP servers with integrations to issue trackers, private docs, and databases
• CLI with interactive commands (interpreter, resume, fork, side), project initialization, and developer docs for connecting models and building harnesses

Use Cases

• Automate large-scale code refactors across a repository using Open Interpreter CLI to read and edit project files, run the full test suite inside a sandbox with explicit approval prompts, and produce a ready-to-commit patch or pull request integrated with your git workflow
• Develop and maintain robust automated tests by having the agent generate unit and integration tests, execute them in an isolated sandbox, diagnose failures, propose and apply fixes with your approval, and push CI-ready changes to accelerate release cycles
• Create up-to-date developer documentation, changelogs and onboarding guides by letting the CLI agent analyze code and README files locally (on-device inference), update docs or examples, run verification commands in a permissioned sandbox, and submit polished PRs while keeping sensitive code off external servers

Who is it for?

• Software developers
• Devops engineers
• Quality assurance engineers
• Machine learning engineers
• Engineering managers