VibeView runs your tests on live cloud phones, tablets and TVs: iPhone, iPad, Android, Apple TV, Android TV and Roku. Bring the flows you already wrote for Maestro: each YAML command becomes a plain-language step, your setups and test data come with it, and importing uses no AI. After a run passes, its learned path is saved and replayed next time, so a repeat run calls the AI agent (beta) only for Verify steps and where the screen changed.
Maestro's open-source framework is free to run on your own machines, and its YAML is a good way to write a flow. Maestro Cloud hosts Android, iOS and web devices to run those flows, priced per device. VibeView takes the same flows further: onto Apple TV, Android TV and Roku, onto a live device you can also use yourself, and into runs that replay without AI.
Import a flow before you read the rest
Upload your .maestro folder as a .zip and click Preview to see every test, its steps and anything that needs a rewrite,
before anything is saved. The free plan includes 30 streaming minutes
and needs no card.
Maestro Cloud alternatives: how VibeView compares
| What you are deciding | VibeView | Maestro Cloud |
|---|---|---|
| Which devices do your flows run on? | Phones, tablets and TVs — iOS simulators (iPhone, iPad), Android emulators, Apple TV, Android TV and physical Roku devices | iOS simulators and virtual Android devices, plus web; its supported-platforms page lists no TV platform |
| What happens to your Maestro YAML? | Imported as readable steps with no AI; setups and test data carry over; a preview flags every command that needs a rewrite (scripts, conditional flows, and touch-only commands on TV) | Runs as written |
| What happens when the screen has changed? | Smart replays the learned path and hands a step it can't replay to the AI agent (beta), which finds the target on the new screen and carries on; No AI replays only; AI only lets the AI run every step | Runs the flow's commands as written; Maestro's pricing lists self-healing with local agents |
| Can you use the device yourself? | Yes — the same live device in your browser, to reproduce a failure, share by link or work through it with a teammate | Not listed — the Cloud plan lists hosted devices for test runs, reporting, CI and PR integration |
| How do you review a run? | A video synced to the steps, the app log, and a screenshot of every step | Step-by-step video playback, detailed logs and flake detection |
| How does it fit your CI and pull requests? | GitHub Actions, any CI through the CLI, checks on the pull request, Slack and webhooks, and cloud builds that compile the app without a Mac | GitHub Actions, any CI through the CLI, results on the pull request, Slack, email and webhooks |
| How is it priced? | Per organization: Team is $49 a month with 4 sessions at once and Business $249 with 10; past the included minutes, streaming is $0.02 a minute. A free plan includes 30 minutes. See every plan | $250 per device a month, priced by the most tests running at once; a free 7-day trial |
The rows that decide it: TV, flows that become readable steps, a learned path that replays without the AI, and pricing per organization instead of per device. Neither runs physical phones: Maestro Cloud's iOS runs are on simulators and its Android runs on virtual devices.
What the import does with your flows
tapOn: "Sign in" becomes Tap "Sign in"; an element's Maestro id stays in the step. assertVisible becomes a Verify step, launchApp with clearState at the start of a flow becomes the
test's start state, and a runFlow at the start becomes a setup that runs before
the test. ${PASSWORD} becomes {{PASSWORD}}, and a value
whose key looks like a password or token is stored as a secret, so the value never appears
in step text.
VibeView never runs scripts from an upload, so JavaScript, conditional flows and a few device commands are flagged for you to rewrite as a step. TV devices have no touch: on Apple TV, Android TV and Roku a point tap, swipe, long-press or double-tap is flagged with a suggested remote step such as "Select …", and Maestro's remote d-pad keys become remote presses. The Maestro import docs list every command, and the guide Bring your Maestro flows to VibeView walks through an import.
After the import
An imported test has no recorded actions yet, so its first Smart run has the AI (beta) carry out each step. When it passes, its learned path is saved and later runs replay it. The AI then only checks the Verify steps and any screen that changed; the guide Tests that stop paying for AI explains how. Or record the flow yourself on a live device. A phone or tablet test recorded on iOS runs on Android too, and a TV test runs on Apple TV and Android TV.
Put tagged groups in CI with vibeview run-suites: CI testing has the GitHub Actions workflow, and schedules run a group every night. For TV apps, the TV testing guide covers the d-pad and focus. Still
choosing a framework? Appium vs Maestro compares
the two, and Appium alternatives weighs
Maestro, XCUITest and Detox.
Frequently asked questions
Can I run my existing Maestro flows on VibeView?
Yes. Upload one flow file or a .zip of your flows folder in the dashboard, or run vibeview import-maestro from the CLI. Each command becomes a plain-language step you can read and edit, and setups ("Before this test") and test data carry over. Before anything is saved, a preview lists every command that needs a rewrite. Importing uses no AI and costs nothing.
What does not carry over from Maestro?
VibeView never runs scripts from an upload, so runScript, evalScript and other JavaScript are flagged for a rewrite, as are conditional flows (when: visible), repeat with while:, clipboard commands and onFlowComplete. On TV apps, touch commands (point taps, swipe, long-press, double-tap) are flagged too, with a suggested remote step such as "Select …". The preview shows each one and where it is in the flow.
Does Maestro Cloud run Apple TV, Android TV or Roku apps?
Maestro's documentation doesn't list any TV platform: on 23 October 2026, its supported-platforms page listed Android, iOS and web, and no TV platform. VibeView runs Apple TV and Android TV on simulators and emulators and Roku channels on physical Roku devices, and imported flows run on them.
Do imported tests use AI on every run?
Only until a run passes. An imported test has no recorded actions, so its first Smart run has the AI (beta) carry out each step. A passing run saves a learned path, and later runs replay that path and call the AI only where the screen changed and for Verify steps. A No AI run never calls the AI; it can run an imported test once a Smart run has saved a learned path. You can also skip the first AI run by recording the test yourself on a live device.
How does the pricing compare?
Maestro Cloud lists $250 per device a month, priced by the most tests you run at once. VibeView prices per organization: Team is $49 a month with 4 sessions at once and 3,000 streaming minutes, and Business is $249 a month with 10 sessions at once. Past the included minutes, streaming is $0.02 a minute. The free plan includes 30 streaming minutes a month and needs no card. Every plan is on the pricing page.
Does either of them use physical phones?
No. Maestro Cloud runs iOS tests on simulators and Android tests on virtual devices; VibeView runs iPhone, iPad, Android, Apple TV and Android TV on simulators and emulators. VibeView's Roku devices are physical hardware.
Published 23 October 2026 · Maestro website and documentation verified 23 October 2026