OpenAI just pushed Codex onto phones. According to TechCrunch AI, the company integrated its coding tool into the ChatGPT app on Thursday, letting developers monitor and manage their coding workflows from iOS and Android devices. The update is in preview and available across all ChatGPT plans.
This isn’t a watered-down mobile version. Codex on phone mirrors what’s running on your desktop, so you can jump in from anywhere your day takes you.
What the mobile Codex actually does
TechCrunch AI reports the new function gives users a live view of Codex environments running on any of their devices. OpenAI summed it up in a statement: “This is more than the ability to remotely control a single task or dispatch new tasks to your computer. From your phone, you can work across all of your threads, review outputs, approve commands, change models, or start something new.”
The key capabilities:
- Cross-device thread access. Pick up any active Codex thread from your phone, no matter where it started.
- Review outputs on the go. See what the agent produced without needing to sit at your machine.
- Approve commands remotely. Codex still waits for human sign-off on sensitive actions, and now that gate works from your pocket.
- Switch models mid-task. Swap the underlying model without restarting the workflow.
- Kick off new tasks. Start fresh work from the phone and let it run on your desktop or in the cloud.
Part of a bigger Codex push
This mobile drop is the latest in a string of upgrades. Last month, OpenAI gave Codex the ability to run in the background in desktop environments, letting it handle tasks autonomously while you work on something else. Earlier this month, the company shipped a Chrome extension that lets the agent operate inside live browser sessions.
String those together and a pattern shows up. OpenAI wants Codex everywhere a developer might be: desktop, browser, phone. The tool used to live in one place. Now it follows you.
The Anthropic angle
What stands out here is the timing. TechCrunch AI notes that Anthropic released a similar feature called Remote Control back in February, which lets users monitor Claude Code’s work from afar. OpenAI is catching up on a capability Anthropic already shipped.
The feature flurry from both sides isn’t an accident. TechCrunch AI frames it as direct competition over whose agentic coding tool wins the developer market. Over the past year, Claude Code has gained real traction among businesses and tech professionals, though both tools remain widely used. Mobile parity matters when developers are picking which agent to standardize on.
Who gets it and when
Availability is straightforward:
- Platforms: iOS and Android, inside the ChatGPT app
- Plans: All ChatGPT tiers, no paywall gate on this preview
- Status: Preview, rolling out now
- Cost: No additional charge beyond the existing ChatGPT plan
The preview label is worth flagging. Features in preview can change, break, or get pulled. Treat this as a real tool to try, not a production dependency yet.
Why this matters
Agentic coding tools are starting to behave less like IDE plugins and more like background workers you check on between meetings. Background mode, browser extensions, and now mobile control all point to the same shift: the developer’s job is moving from typing every line to supervising agents that type for them.
If you’re already using Codex, install the latest ChatGPT mobile app and check whether the preview is live on your account. If you’re on Claude Code, Anthropic’s Remote Control has had a three-month head start, so the workflow is already mature there.
The next question is which company ships the next capability first. Full details on the rollout are available at the original TechCrunch AI report.