Skip to content
Let's Talk
Claude Code

Claude Code Remote Control: Control Your AI Coding Agent from Your Smartphone

Felix Schmidt

On February 25, 2026, Anthropic released one of the most requested features for Claude Code: Remote Control. The idea is as simple as it is powerful – you start a coding session at your desk and continue controlling it on the go from your phone.


What Is Remote Control?

Remote Control lets you monitor and control a running Claude Code session from a smartphone, tablet, or any web browser. The key difference from cloud-based solutions: your code stays on your machine. The phone acts purely as a remote control – similar to tools like Tailscale.

As Anthropic puts it: "Unlike Claude Code on the web, which runs on cloud infrastructure, Remote Control sessions run directly on your machine and interact with your local filesystem. The web and mobile interfaces are just a window into that local session."


How It Works

Setup is straightforward:

  1. Type /remote-control or /rc into your terminal
  2. Claude Code automatically generates a QR code and a session URL
  3. Scan the QR code with the Claude app (iOS/Android), or open the URL in any browser
  4. Alternatively, select the session from the list at claude.ai/code
  5. The live session instantly appears on your phone, fully in sync

All MCP servers, skills, and project settings are fully preserved. The conversation stays in sync in real time across all connected devices.


Security by Design

Anthropic put a strong emphasis on security in the design:

  • No inbound ports are opened on your machine
  • The connection runs exclusively via outbound HTTPS requests
  • Each access token is valid for a single session only – the link becomes invalid automatically after the session ends
  • All traffic travels through the Anthropic API over TLS
  • No port forwarding or complex VPN configuration required

Availability

Remote Control is available from version 2.1.52 of Claude Code. The feature is available on all plans (Pro and Max) – it launched as a research preview for Max subscribers first, with Pro users following. On Team and Enterprise plans, Remote Control is off by default and must be enabled by an admin at claude.ai/admin-settings/claude-code.


FAQ: The Most Important Questions

Do I need to keep my computer on the entire time?

Yes – the terminal must remain open. Remote Control runs locally on your machine. If the terminal is closed, the remote session ends. The machine doesn't have to be actively in use, but the process must be running in the background.

What happens if the laptop goes to sleep / enters energy saver mode?

The session is sleep-resilient: if the laptop briefly enters sleep mode or the network drops momentarily, the session automatically reconnects once the machine comes back online. However, a network outage of around 10 minutes or more will cause the session to time out and require a restart.

Do I need a static IP address?

No. Since Remote Control exclusively makes outbound HTTPS connections to the Anthropic API, no static IP, port forwarding, or special network configuration is required. The feature works behind firewalls and VPNs without any additional setup.

Can I connect multiple devices simultaneously?

No – only one remote connection per session is supported at a time. However, multiple Claude Code instances on the same machine can each run independent sessions with their own remote connections.

Can I start a new session directly from my phone?

Not yet. Remote Control requires the session to be started on the computer first. If you want to start a fresh session on a specific project while away from your desk, you need to initiate it at the desktop first. Fully mobile-initiated sessions are not yet supported.

Does Remote Control work with API keys?

No. Remote Control requires claude.ai OAuth authentication (/login → choose the claude.ai option). API keys are not supported for this feature. If ANTHROPIC_API_KEY is set as an environment variable, it must be unset first.

Is the feature available for teams and enterprises?

On Team and Enterprise plans, Remote Control is disabled by default. Admins can enable it at claude.ai/admin-settings/claude-code. For organizations with certain data retention or compliance configurations, the toggle may be grayed out – in that case, Anthropic recommends contacting support directly.


Why Does This Matter?

Remote Control addresses a real pain point in developer workflows: long autonomous agent sessions that you don't want to sit and watch at your desk. With this feature, you can step away, take a coffee break, or check progress on your commute – without interrupting the session or resorting to insecure remote desktop solutions.

Combined with the new /loop command (which repeats tasks automatically) and the 1-million-token context window for Opus 4.6, Claude Code is increasingly becoming a true autonomous developer agent – one you can set in motion and accompany from a distance.


Sources

Official Anthropic Documentation

This topic relevant to your team? Let's discuss how I can help.

This website uses third-party services (Google reCAPTCHA, Calendly) that may set cookies. See our Privacy Policy for details.