Does CodeAgent Mobile work with the OpenAI Codex CLI?

Yes. codeam-cli wraps Codex under a pseudo-terminal, so you launch it with `codeam codex` instead of `codex` and every interactive flow — approvals, plan steps, file prompts — is forwarded to your phone exactly as Codex emits it. The rest of your setup is unchanged.

Do I need a separate OpenAI or ChatGPT subscription?

Yes. CodeAgent Mobile is a relay, not an LLM provider. Codex runs on your own machine with your existing ChatGPT plan or OpenAI API key; CodeAgent only meters the relay between your phone and your terminal.

Does my code leave my computer?

No. Codex runs locally and reads and writes files on your machine. The CodeAgent backend only relays prompts and responses between your phone and the terminal — it never sees your repository.

How do I run Codex instead of Claude Code?

Start it explicitly with `codeam codex`, or pair a Codex-specific session with `codeam pair --agent codex`. `codeam sessions switch` lets you move between paired machines and agents at any time.

Can I use it inside SSH sessions, Docker, or a Codespace?

Yes. codeam-cli only needs Node.js 18 or newer and a network connection, so it runs inside SSH sessions, tmux/screen, Docker containers, and GitHub Codespaces — anywhere you can start a Node.js process.

Is there a free tier?

Yes — CodeAgent Mobile has a free plan with no credit card: a daily task allowance, up to 2 paired devices, and 1 linked agent. Pro removes the daily caps and adds Team Spaces, the self-hosted deploy plane, Agent Toolkit integrations, and built-in cost-saving compression. See the pricing page at codeagent-mobile.com/pricing for current Pro pricing.

For OpenAI Codex users

Control OpenAI Codex from your phone

CodeAgent Mobile is a mobile companion app that lets you control OpenAI Codex from your phone. Install codeam-cli, pair your terminal with the app once, then run `codeam codex` and send prompts, stream responses, and approve interactive prompts directly from iOS or Android while Codex runs in your local terminal with your existing OpenAI or ChatGPT credentials.

Download for iOS Download for Android Install codeam-cli

How it works

Pair your phone with OpenAI Codex once. After that, every prompt and response flows between the two automatically.

1
Run `npm install -g codeam-cli` in your terminal (requires Node.js 18 or newer, plus the OpenAI Codex CLI installed and signed in with `codex login`).
2
Run `codeam pair` to generate a 6-character pairing code and a QR code in your terminal.
3
Open CodeAgent Mobile on your phone, tap "Pair Device", and scan the QR or enter the 6-character code.
4
Run `codeam codex` — your phone is now the remote. Send prompts, watch responses stream live, and approve any prompts Codex surfaces.

npm install -g codeam-cli

Why developers use it

A real remote control for your AI coding agent — not a chat fallback.

Real-time streaming

Token-by-token output streams from your laptop straight to your phone — diffs, file lists, and the agent's reasoning all render live.

Secure end-to-end relay

The CodeAgent backend is a thin relay. Your code never leaves your machine, and the AI agent runs with your existing API key or subscription.

Works with your existing setup

No changes to your IDE settings, Anthropic key, GitHub Copilot subscription, or Cursor account. Pair once and start sending prompts.

Approve interactive prompts

When the agent stops to ask y/n or pick an option, the prompt appears on your phone. Tap to approve, reject, or pick — without going back to the keyboard.

Frequently asked questions

Specific to OpenAI Codex users.

Get started in under a minute

Install the OpenAI Codex client, pair your phone, and start sending prompts from anywhere.

Install codeam-cli npm install -g codeam-cli