Anthropic Extends Claude Code to Mobile with ‘Remote Control’ Feature
Claude Code, Anthropic’s AI-powered coding assistant, has rapidly gained traction since its launch, attracting both developers and non-technical users seeking to build applications and websites with unprecedented speed. The platform, which leverages plain English to generate code, has been primarily accessible through desktop applications and terminal interfaces – until now. Anthropic has introduced Remote Control, a new mode that extends Claude Code’s functionality to iPhone and Android smartphones, initially available to subscribers of the Claude Max tier ($100-$200 USD monthly).
Anthropic announced on X that Remote Control will eventually be rolled out to users on the Claude Pro ($20 USD monthly) subscription level as well.
A Mobile Command Center
According to Claude Code Product Manager Noah Zweben, Remote Control establishes a synchronization layer that connects local command-line interfaces with the Claude mobile app and web interface. This allows developers to initiate complex tasks in their terminal and maintain control from a phone or tablet, effectively decoupling the AI agent from the physical workstation.
Currently in Research Preview for Claude Max subscribers, Remote Control is positioned as a tool for power users. Access is not yet available for Team or Enterprise plans. To activate the feature, users must update to Claude version 2.1.52 and execute the command claude remote-control or use the in-session slash command /rc. A QR code is then displayed in the terminal, which, when scanned, opens a synchronized session within the Claude mobile app. Detailed setup instructions are available on the Claude Code documentation site.
Preserving Flow State
The launch of Remote Control is framed by Anthropic as a means of enhancing developer productivity and preserving “flow state.” Zweben emphasized the lifestyle benefits, encouraging users to “take a walk, see the sun, walk your dog without losing your flow.”
The core principle behind Remote Control is that it doesn’t replace local development environments but rather provides a remote portal into them. As outlined in official documentation, Claude continues to run on the user’s machine, maintaining access to the filesystem, environment variables, and Model Context Protocol (MCP) servers.
Architecture and Security
Claude Code Remote Control operates as a secure bridge between a local terminal and Anthropic’s cloud interface, which hosts the Opus 4.6 and Sonnet 4.6 AI models powering Claude Code. The connection is initiated by the user’s desktop machine, establishing an outbound connection to Anthropic’s API. This design avoids the need to open inbound ports or expose the user’s computer to the public internet. Instead, the local machine polls the API for instructions.
Using the mobile app or accessing the session URL essentially provides a remote viewing and command interface for the process running on the user’s computer. Files and MCP servers remain local; only chat messages and tool results are transmitted through the encrypted connection.
Users must be on a Pro or Max plan and have authenticated their CLI using the /login command to begin. Running claude remote-control in the project directory generates a unique session URL and a QR code (toggleable with the spacebar) for mobile access.
From Workarounds to Official Support
Prior to the official release of Remote Control, developers relied on unofficial workarounds to access terminal-based workflows from their mobile devices. These solutions often involved a combination of tools like Tailscale for secure tunneling, Termius or Termux for mobile SSH access, and Tmux for session persistence. Some developers even created custom WebSocket bridges to achieve a responsive mobile UI for their Claude sessions.
These unofficial methods, while functional, were often unstable and prone to timeout issues. Remote Control offers a native streaming connection that eliminates the need for port forwarding or complex VPN configurations, and includes automatic reconnection logic to maintain sessions even if the host machine sleeps or loses network connectivity.
A $2.5 Billion Agent
The launch of Remote Control represents a significant step forward for Anthropic’s Claude Code, which has become a major revenue driver for the company. As of February 2026, Claude Code has reached a $2.5 billion annualized run rate, more than doubling its revenue since the beginning of the year.
Claude Code is currently experiencing rapid adoption, with 29 million daily installs within Visual Studio Code. Analysis suggests that 4% of all public GitHub commits are now attributed to Claude Code.
By extending its capabilities to mobile, Anthropic is solidifying its position in the “agentic” coding space, moving beyond simple code completion to a model where AI functions as an autonomous collaborator.
Looking Ahead: Vibe Coding Everywhere
The move to mobile terminal control signals a broader shift in the software development landscape. The industry is entering an era where AI tools are responsible for approximately 41% of all code written. For developers, this translates to a shift from line-by-line coding to strategic oversight.
This trend is expected to accelerate as mobile-tethered agents become more prevalent. The decreasing barrier between “idea” and “production” could enable single developers to manage complex systems that previously required entire DevOps teams. This shift has already impacted the tech market, with shares of cybersecurity firms like CrowdStrike and Datadog falling as much as 11% following the launch of Claude Code’s automated security scanning features.
As Claude Code transitions from the desktop to the pocket, the role of the software engineer is being redefined. The coming year may see a rise in “one-person unicorns”—startups built and maintained largely through mobile agentic commands—potentially marking the end of traditional manual coding.
