OpenAI brings Codex remote coding controls to phones
OpenAI has added Codex controls to the ChatGPT app on iOS and Android, letting users monitor and manage coding workflows from a phone. The preview feature is available across all plans, and OpenAI said it lets users review outputs, approve commands, change models, and start new tasks remotely.
OpenAI is bringing Codex to mobile. The company said Thursday that the coding tool is now integrated into the ChatGPT app, giving users a way to monitor and manage development workflows from phones and other mobile devices.
The update is in preview and is available to all plans on iOS and Android. According to OpenAI, users can see Codex live environments on any device where the tool is running, which means they can keep track of work that is happening on a desktop or server without being tied to that machine.
⚡ New to this?
This news is about OpenAI adding mobile controls for Codex, its coding agent. A coding agent is software that can write or modify code and carry out tasks with some independence, while the user reviews and approves the work.
For non-experts, the big deal is that developers can now watch and manage that work from a phone instead of sitting at a computer. That matters because these tools are becoming part of normal software workflows, not just experiments in a laptop browser.
🦞 OpenClaw angle
If you run self-hosted AI coding agents, think about the control plane, not just the model. Build mobile-friendly monitoring for task status, output review, and approval flows so operators can pause or approve work without opening a laptop.
Also separate “background execution” from “human approval” in your automation design. Keep commands, model changes, and task starts behind explicit approval steps, since OpenAI is highlighting those controls as part of the product direction.
OpenAI described the mobile feature as more than simple remote control. In a statement, the company said users can “work across all of your threads, review outputs, approve commands, change models, or start something new” from their phone.
The move follows a series of recent product updates around Codex. Last month, OpenAI said Codex could run in the background in desktop environments, letting it carry out tasks on its own. Earlier this month, the company also introduced a Chrome extension so the agent can work inside live browser sessions.
The timing also shows how aggressively OpenAI and Anthropic are competing in agentic coding tools. In February, Anthropic released a similar feature called Remote Control, which lets users monitor Claude Code from afar.
That competition has intensified over the past year as both companies push coding agents that can do more than answer prompts. OpenAI and Anthropic are both trying to make their tools useful for developers and other technical users who want software that can handle parts of a workflow on its own.
According to the source article, Anthropic’s Claude Code has gained popularity among businesses and tech professionals over the past year, even though both tools are still widely used. OpenAI’s latest Codex update is another step in the race to make agentic coding tools easier to supervise and use outside a desktop setup.
