OpenAI Brings Codex to ChatGPT Mobile for All Plans, Chasing Claude Code
Updated
Updated · The Verge · May 14
OpenAI Brings Codex to ChatGPT Mobile for All Plans, Chasing Claude Code
19 articles · Updated · The Verge · May 14
OpenAI is rolling out a preview that lets ChatGPT users on iOS and Android control Codex running on their computer from a phone, including starting tasks, approving commands and switching models.
Real-time updates flow back to the mobile app with screenshots, terminal output, diffs and test results, while files, credentials, permissions and local setup remain on the machine where Codex operates.
The feature is available across all ChatGPT tiers, including the free plan and the lower-cost Go plan, widening access beyond desktop use.
OpenAI’s move follows rising demand for Anthropic’s Claude Code and builds on a recent Codex update that let the tool operate apps on macOS as part of a broader desktop superapp push.
With AI agents gaining desktop control from a phone, have we created an unstoppable new cybersecurity threat?
Amidst fierce competition, can OpenAI's new AI agent solve its 'mile wide, inch deep' user engagement problem?
As AI agents replace human browsing, how must websites evolve to survive in a machine-first world?
OpenAI Codex Hits Mobile: The $25B Race for Agentic Coding and Workflow Automation
Overview
OpenAI has expanded its Codex agent to all ChatGPT plans, including mobile, aiming to make Codex accessible to a wider audience beyond just developers. By introducing 'codex remote-control,' users can now manage Codex sessions directly from their mobile devices, integrating powerful AI capabilities into everyday digital workflows. This move is supported by ongoing efforts to ensure secure and stable remote session management, with a focus on proper authorization and security best practices. The expansion marks a strategic step for OpenAI, positioning Codex as a versatile tool for both technical and non-technical users in their daily tasks.