Expo Opens Paid-Plan Access to Cloud Simulators for AI Coding Agents

By Amara Osei

Expo opened Expo Simulators to every paid plan on November 12, 2026. The company priced cloud iOS simulator and Android emulator usage at $0.10 per simulator minute on either platform.

The launch expands a limited-access EAS Simulator preview that Expo introduced on September 14, 2026. An Expo blog post dated November 10 still directed developers to a waitlist, making paid-plan access a broader release rather than the service’s first public demonstration.

Where does the verification bottleneck sit?

Expo says code generation has shifted the bottleneck to verification because agents produce changes faster than engineering teams can test and review them. The company identifies a mobile constraint: iOS simulators run only on Apple hardware, while cloud agents typically run on Linux. Without a runtime, an agent must assess code without seeing app behavior, according to Expo.

Running locally also consumes developer capacity. In Expo’s example, seven parallel agents can require seven simulators or emulators, and each Android emulator can use 4 GB or more of RAM. Expo says remote sessions move that compute to its infrastructure and allow an iOS simulator to run from Linux, Windows through the Windows Subsystem for Linux, or a cloud sandbox without a local Mac.

How does Expo Simulators fit into an agent workflow?

Expo’s documented workflow starts with a bug report in a tool such as Linear. According to Expo, an agent reproduces the issue, writes a fix and tests it on cloud devices in parallel, away from the developer’s machine. Expo says the agent can attach a screen recording, screenshots and logs to a pull request, while teammates open the running build from a link without a TestFlight build.

Expo says the service works with Cursor, Claude Code and Codex. It adds no AI model of its own; the company supplies the device runtime and reviewable evidence instead of another coding model.

Expo co-founder and CTO James Ide said: “The agent was the fast part. Expo Simulators gives every agent a simulator, so a fix arrives for review already tested on several simulators, and the only decision left is whether it’s the right code. Watch it run, then merge.”

What can a cloud session run?

The Expo Simulators documentation says users can start a session through EAS CLI, the REST API or an authenticated creation link. A session can install an Expo Build, a build whose native code matches the project’s fingerprint, an application archive or Expo Go; without an application source, it starts with a blank device.

Expo says agents and test jobs can control apps through agent-device, Argent or Appium, while a browser preview lets people watch the run. The official Expo Simulators skill packages access checks, build preparation, session limits, device control, evidence collection and cleanup for agents.

The product page lists support for React Native, Swift, Kotlin, Flutter, Unity and other frameworks. Expo says apps need not use Expo or React Native, but the build must run on an iOS Simulator or Android Emulator and the project must link to Expo Services.

Expo says each session runs in an isolated environment with session-scoped credentials and access-controlled previews. Stopping a session disables the simulator and controller endpoints, according to the company.

What do access and capacity cost?

Expo’s pricing page lists 60 simulator minutes per month on the Starter, Production and Enterprise plans until 2027, with usage-based pricing beyond that allowance. Paid sessions can run for up to 115 minutes.

Parallel use does not create unlimited capacity, according to Expo’s usage and billing guide. The guide says simulator sessions draw on the account’s job concurrency and queue when no capacity is free. Expo says this makes cleanup and session limits part of cost control for automated workflows.

Ending a CI job or returning from a non-interactive command does not stop the remote device, according to Expo. The company requires teams to stop sessions explicitly or set a duration or idle limit; otherwise, an unattended session keeps consuming billable usage.

Where does cloud verification stop?

Expo’s limitations guide says the simulator CLI commands remain experimental and the service provides no physical devices. The guide also warns that recordings may miss frames, so they cannot prove or disprove a brief frame drop. Expo says a cloud runtime can verify app behavior, but it does not replace physical-device testing or profiling where those tools are required.

What comes next?

Expo’s next disclosed step is Agent Sessions, a planned enterprise layer for triggering agents from customer feedback and protecting against prompt injection. The company says malicious input could otherwise direct an agent to expose data, and it expects to share Agent Sessions details by the end of November 2026.

Expo says Agent Sessions remains a roadmap item rather than a capability in the current launch. According to Expo, Simulators adds a verification stage to the company’s existing build, CI/CD, app-store submission, over-the-air update and monitoring services. Expo frames the result as a division of labor rather than an autonomous release process: agents provide tested changes and runtime evidence, account limits govern concurrent runs, and people retain the merge decision. Expo retains that merge authority in the current launch while expecting to disclose its enterprise Agent Sessions plans by the end of November 2026.