A terminal-based coding agent that runs locally on the user's computer and requires an external OpenAI ChatGPT account or API key for model inference. It is installed via package managers like npm or Homebrew and interacts through a command-line interface.
Project overview
The project addresses the need for a lightweight coding agent that runs locally in the terminal while integrating with OpenAI models via a ChatGPT plan or API key.
Project type
AI Agent · AI Coding
Use cases
Coding & Development
Deployment
Refer to project documentation
License
Apache-2.0
Best for
Developers seeking a lightweight coding agent that executes locally in the terminal and connects to OpenAI models.
Key capabilities
A coding agent that runs locally on the user's computer in the terminal
Allows users to sign in with their ChatGPT Plus, Pro, Business, Edu, or Enterprise plan to use the agent
Limitations and risks
A ChatGPT account or OpenAI API key is required, meaning the data boundary includes required external services and remote inference; local execution does not provide a local-only data boundary.
Usage may depend on a paid service depending on the chosen OpenAI account or API tier.
GPU and minimum hardware specifications are not documented.
Getting started
Installation difficulty is rated as easy because it offers package manager installation via npm or Homebrew. The first success path is to install via package manager or script, then run the codex command.
Evidence and sources
GitHub project description: Lightweight coding agent that runs in your terminal
README: Run the following on Mac or Linux to install Codex CLI
README: npm install -g @openai/codex
README: brew install --cask codex
README: We recommend signing into your ChatGPT account to use Codex as part of your Plus, Pro, Business, Edu, or Enterprise plan.