OpenAI adds Codex to mobile ChatGPT so developers can manage coding tasks from anywhere

OpenAI has expanded access to Codex and built it into the ChatGPT mobile app. Developers can now monitor coding tasks, steer them and approve actions in real time, even when they're away from their laptops.

Codex on your phone

The core idea behind the announcement is simple: Codex is no longer tied to the desktop. If a task is already running on another device or in a remote environment, users can open ChatGPT on their smartphone to quickly check its status, review progress and step in if needed. That's especially useful when code is being generated, tests are running and the next step needs a human sign-off. Mobile access turns Codex from a tool for sessions at the computer into an always-on work channel.

According to OpenAI, this goes beyond checking status. Through mobile ChatGPT, developers can give follow-up instructions, steer tasks and approve actions as needed. In effect, the company is betting on a workflow where developers stay in the decision loop without having to sit in front of a screen the whole time. The format fits best when the work already runs in a remote environment or on a separate device.

How the workflow looks

The new mode is a logical extension of asynchronous development with an AI assistant. You kick off a task wherever it's convenient to work with the repository and environment, then follow it from your phone while you walk to a meeting, ride across town or just step away from your desk. If Codex hits a fork in the road, asks for confirmation or shows an intermediate result, you don't have to go back to your laptop for a single click.

In this setup, the mobile app covers several practical use cases:

• Checking task status and progress
• Quickly approving actions that require human input
• Course-correcting when a request needs clarifying or a plan needs changing
• Monitoring tasks launched on another device or in a remote environment

"Monitor, steer and approve coding tasks in real time."

This isn't an attempt to replace a full IDE with a smartphone. It's closer to a mobile control panel for moments when work is already underway and the job isn't writing code from scratch but overseeing the process. The approach fits the modern stack well: compute, agents and environments increasingly live apart from the developer's local machine, while decisions still need to be made without being tied to a single screen.

Why it's convenient

For OpenAI, this is another step toward a model where Codex isn't a one-off feature but a persistent agent working alongside the developer. The less friction there is between the task, the device and the moment of decision, the easier it becomes to fit AI into a real workday. Many workflows used to stall the moment a person walked away from the computer. Now the process gains continuity: the task keeps running, and the user stays connected.

For teams, it's a practical change too. When tasks run in remote environments, on servers or on another device, mobile access cuts approval delays and unblocks the next step faster. There's no need to wait for someone to get back to their desk to approve an action or tweak an instruction. If OpenAI makes this workflow stable enough for everyday use, the line between "I'm coding" and "I'm managing the coding" will get even thinner.

What it means

OpenAI is pushing Codex toward being an always-available assistant rather than a tool limited to desktop sessions. For developers, it's a signal that managing AI tasks is gradually becoming cross-platform and asynchronous: you can start some of the work in one place and monitor and approve it in another without losing momentum. If this mode catches on, mobile ChatGPT will become a working interface not just for conversation but for day-to-day development oversight.