How to Use OpenAI Codex from Your Phone: The 2026 iOS and Android Guide

OpenAI shipped Codex on mobile this week. On May 14, 2026, the ChatGPT app on iOS and Android added a full Codex experience for every plan, including Free and Go. You can now monitor running tasks, approve commands, swap models, and start new work from your phone while your laptop is idle.

Try Apidog today

Apidog readers have been asking when serious AI coding would reach a phone. The answer is now. This guide covers what shipped, how to enable it, and the first workflow worth testing.

For terminal-first workflows, start with Apidog's Codex CLI setup guide. If you want to compare other mobile agent setups, read the Claude Code on mobile walkthrough and Run Cursor on your phone. You can also use Apidog to build and test APIs that phone-side Codex tasks can call.

What "Codex from anywhere" means

OpenAI's announcement covers four surfaces:

1. Codex in the ChatGPT mobile app: iOS and Android, preview, all plans
2. Codex in Slack: Plus, Pro, Business, Enterprise, and Edu; mention @Codex in a thread
3. Codex Chrome extension: shipped May 7, 2026; works across tabs without taking over the browser
4. Codex SDK: programmatic control of Codex from scripts and CI

Mobile is the most visible change, but Slack and the SDK are what make this useful for teams. A senior engineer can assign work to Codex from a Slack thread, then approve the result from a phone later.

Set up Codex on iOS and Android

The mobile experience lives inside the existing ChatGPT app. You do not need a separate Codex app.

Step 1: Update ChatGPT

Install the latest ChatGPT app from the App Store or Play Store.

Mobile Codex requires the May 13, 2026 release or newer, according to the Codex changelog.

Step 2: Sign in with your desktop account

Use the same OpenAI account you use for ChatGPT or Codex CLI.

Mobile Codex uses the same threads, environments, and connected hosts as the web app.

Step 3: Connect a cloud environment

If you only use Codex from your terminal, configure at least one cloud environment first.

In the web app:

1. Open Settings
2. Go to Codex
3. Open Environments
4. Link GitHub
5. Configure a repository

The mobile app inherits this setup.

Step 4: Open the Codex tab

In the ChatGPT mobile app, tap Codex in the bottom navigation.

You should see active tasks and existing threads.

Step 5: Run a small test task

Start with a low-risk change:

Add a short docstring to the parseConfig function.

Or:

Update the README with one sentence explaining how to run tests.

Then verify the full loop:

1. Codex creates the change
2. The diff renders correctly on your phone
3. You review the change
4. You tap Approve

If that works, you can move on to larger tasks such as multi-file refactors or issue-driven fixes.

What you can do from your phone

The mobile app is best used as a control surface for Codex. From your phone, you can:

• Watch live runs across laptops, devboxes, or connected remote environments
• Browse threads and switch between parallel tasks
• Review diffs before they merge into a branch
• Approve commands Codex wants to run
• Swap models when a task needs a stronger model
• Start new tasks from a prompt or GitHub issue
• Comment on a pull request opened by Codex

OpenAI describes it this way: "From your phone, you can work across all of your threads, review outputs, approve commands, change models, or start something new."

The limitation: mobile is not a full code editor. You are not editing raw code like you would in VS Code. Codex writes the code; you guide and approve the work.

Use Slack to hand work to Codex from a team thread

Slack support makes Codex feel closer to a teammate.

How it works

After a workspace admin installs the Codex Slack app from the Marketplace, team members can mention @Codex in a channel or thread with a task.

Codex then:

• Selects the best configured environment
• Defaults to the first repository in the environment map unless you specify another repo
• Reacts with an emoji
• Posts a task link
• Runs the work
• Replies in the thread when the run finishes

Example prompt:

@Codex please investigate the failing checkout test in the payments repo and open a PR with the fix.

If you need to target a specific repo, include it explicitly:

@Codex in acme/payments-api, add a regression test for the checkout timeout bug described above.

Requirements

You need:

• ChatGPT Plus, Pro, Business, Enterprise, or Edu
• A connected GitHub account
• At least one configured cloud environment
• Workspace admin approval to install the Slack app

Free is excluded for Slack.

Enterprise admins can disable answer posting and force Codex to share only task links. This keeps generated code out of channel transcripts. The setting lives in workspace settings.

Useful Slack workflow

A practical pattern is issue triage:

1. A GitHub issue appears in a Slack channel
2. Someone mentions @Codex in the thread
3. Codex attempts a fix
4. The team reviews the linked task or PR
5. A human approves or rejects the change

The OpenClaw GitHub triage bot uses a similar operational pattern.

Use the Codex SDK for automation

The SDK is for scripts, schedulers, and CI workflows. It lets your own tooling drive Codex with your workspace identity.

A common flow looks like this:

from openai import Codex

client = Codex()

task = client.tasks.create(
    repo="apidog/awesome-api",
    prompt="Add OpenAPI examples to every endpoint missing them.",
    environment="prod-mirror",
)

for event in client.tasks.stream(task.id):
    print(event.summary)

You could wire this into:

• A nightly job that opens PRs for stale issues
• A CI step that asks Codex to add a missing test
• A scheduled documentation cleanup
• A repository maintenance workflow

Enterprise workspaces can mint access tokens for non-interactive flows. That capability shipped May 5, 2026.

If your team already runs Claude Code with GitHub Actions, the Codex SDK fills a similar role on the OpenAI side.

Plans, pricing, and availability

Mobile preview is available on every plan, including Free and Go. Other surfaces have tier restrictions.

Surface	Free	Go	Plus	Pro	Business	Enterprise / Edu
Mobile (iOS + Android)	Yes (preview)	Yes	Yes	Yes	Yes	Yes
Slack integration	No	No	Yes	Yes	Yes	Yes
Chrome extension	Yes (preview)	Yes	Yes	Yes	Yes	Yes
Codex SDK	Limited	Limited	Yes	Yes	Yes	Yes
Enterprise access tokens	No	No	No	No	No	Yes

For a deeper cost breakdown, read the GPT-5.5 pricing breakdown. If you want to try Codex without paying, the free Codex for open source guide explains the application route.

How phone Codex compares to alternatives

Mobile coding agents are still a young category. The main options:

• OpenAI Codex: strongest mobile UX, tight ChatGPT integration, available on Free
• Claude Code on mobile: the Claude Code mobile setup uses tmux and SSH; more DIY, but powerful for long-running terminal work
• Cursor on your phone: the Cursor mobile workflow uses remote development plus Cursor's web preview; useful if Cursor is already your desktop IDE

For desktop-side comparison before choosing a mobile workflow, read the Claude Code vs Codex 2026 breakdown. For a broader comparison, see the Copilot vs Claude vs Cursor vs Codex comparison.

Where API testing fits

A phone-based coding agent is useful only if the code it ships behaves correctly. API changes are often where regressions appear.

Apidog gives you an API client, OpenAPI editor, and automated test runner that can run locally or in CI.

A practical workflow:

1. Codex opens a PR that changes an endpoint.
2. Apidog's CI workflow runs the existing OpenAPI test suite against the PR's preview deploy.
3. If the suite passes, you review and approve the diff from your phone.

For implementation details, read the Apidog ChatGPT API testing guide and How to test AI agents that call your APIs. You can also download Apidog to try the workflow.

Common questions

Does Codex on mobile work offline?

No. Codex runs against OpenAI's cloud or your connected environment. Without a network, the app can show your last-known thread state, but it cannot start new tasks.

Can I edit code directly in the mobile app?

Not like a normal editor. You can prompt, review, and approve. The mobile app is a control surface for a remote agent, not a mobile VS Code replacement.

Is the mobile version slower than desktop?

The agent runs on the same backend. What feels slower is the phone UI. Long diffs are harder to scan on a small screen, so a good pattern is:

1. Read the task summary on mobile
2. Approve small, low-risk changes from the phone
3. Move to desktop for large or sensitive reviews

Does mobile Codex support voice input?

Yes, through the existing ChatGPT voice mode. You can dictate Codex prompts the same way you dictate other ChatGPT messages.

What happens if I lose signal during approval?

The task keeps running in the cloud. When your phone reconnects, the state updates. Codex does not require an open mobile session to continue working.

Can my enterprise admin disable mobile Codex?

Yes. Workspace owners can restrict Codex access from the admin panel. The same controls that gate desktop access apply to mobile.

Will Codex on mobile cost extra?

There is no separate fee for the mobile app. You pay for the underlying Codex compute based on your plan. The Codex pricing post has the breakdown.

Is this different from the older "Codex" model?

Yes. The current Codex is a coding agent product, not the deprecated 2021 model. If you still think of Codex as the old API endpoint, start with the Codex CLI introduction.

Try it tonight

Run the smallest complete workflow:

1. Update the ChatGPT app.
2. Sign in with your OpenAI account.
3. Link a cloud environment.
4. Ask Codex to make a one-line README change.
5. Review the diff on your phone.
6. Approve the change.

After that works, add Slack for team handoffs and the SDK for automation.

If you want API contracts to stay correct while Codex ships from your phone, pair it with Apidog. Codex writes the code; your API test suite catches regressions.