Momentic

Visit Website

Introduction

Momentic is a hosted service for end-to-end tests written in plain language or YAML. An AI step can find an element from a description (“the save button”) and act on it. A normal step is cheaper and more predictable. Tests run in Momentic’s browsers, with optional Android emulators and iOS simulators on paid usage. Results, screenshots, and traces stay in the Momentic project.

The product is closed source. Pricing is credits, not seats. A normal step costs 1 credit. A step generated by an AI action or by failure recovery costs 2. Interactive runs in the editor are free. Momentic is a test runner you rent. It is not a library you pin in package.json and execute only on your own CI agents, though CI can trigger the hosted runs.

Features

1. Plain-language and YAML tests: A test is a sequence of steps. AI steps describe the target. YAML steps name the action more tightly.
2. Hosted browsers: The free plan includes 5 concurrent browsers. Pay-as-you-go raises that to 25. You do not install Playwright on a runner to get the first result, because execution is in Momentic’s cloud.
3. Editor: Tests are authored and debugged in Momentic’s UI. Editor runs do not consume credits.
4. Failure recovery: An AI recovery step can try to continue a broken flow. It costs 2 credits and can also hide a real product bug if you treat recovery as success.
5. Mobile sessions: Virtual Android and iOS devices. Free includes one concurrent mobile session, 15 minutes long. Pay-as-you-go includes 5 Android emulators and 5 iOS simulators, still 15 minutes. These are virtual devices, not a hardware lab.
6. Email and SMS: Disposable inboxes are unlimited on the listed plans. Pay-as-you-go adds 5 phone numbers for SMS and one-time passwords. The free plan lists zero phone numbers.
7. CI triggers: Scheduled runs and pipeline hooks start the same hosted tests. The gate is Momentic’s result, so CI credentials and the Momentic project both have to be up.
8. Retention: Full results and artifacts are kept 30 days on Free and Pay-as-you-go. Aggregate analytics are listed as kept indefinitely. Enterprise retention is custom.

Pros

1. Authoring is fast for a UI flow: Describing a button beats writing a selector, until the description matches the wrong node.
2. No seat fee: Credits are the meter. A large team does not buy a license per person on the published plans.
3. Editor runs are free: Debugging a test does not burn the monthly allowance. That is the difference between a usable trial and a frustrating one.
4. Inboxes for signup flows: Email assertions are included, which many UI runners leave as your problem.
5. Free allowance is non-zero: 2,000 credits a month is about 200 runs of a 10-step test, on Momentic’s own estimate, before hosting and AI surcharges.

Cons

1. Credits move under AI: Recovery and AI actions cost 2 credits. A suite that “heals” itself can double the bill and mask breakage.
2. Hard stop on Free: At 2,000 credits the month ends. Unused credits do not roll over on any listed plan.
3. Hosted only: Source of the runner is not yours. Enterprise is the tier that discusses custom security terms.
4. Virtual mobile, short sessions: 15 minutes and emulators or simulators will not catch device-lab issues. Firebase Test Lab or a hardware farm is a different tool.
5. 30-day artifacts: A failure from last quarter is gone on the self-serve plans. Analytics without the screenshot is a weaker record.

Usage and Links (Example)

Create a project at momentic.ai, write a test in the editor, and run it there first (no credits). A step list for a public page looks like:

- action: navigate
  url: https://example.com
- action: click
  description: the primary call to action
- assert: text "Example Domain" is visible

The exact YAML keys are in Momentic’s editor and docs; treat the block above as the shape, and confirm property names in the product before you commit a suite. Connect CI with the project token Momentic issues, and fail the job when the hosted run fails. Start with normal steps. Add AI actions only where a stable selector or description is not available.

Resources:

• Momentic website
• Pricing

Pricing

From momentic.ai/pricing. A normal step is 1 credit. An AI action or failure-recovery step is 2. Editor runs are free. Credits reset monthly and do not roll over.

1. Free: $0. 2,000 credits a month, about 200 ten-step runs before AI and hosting. 5 concurrent browsers, 1 mobile session, 15-minute mobile sessions, unlimited email inboxes, no phone numbers, 30-day results. Usage stops at the cap.
2. Pay-as-you-go: $125 per month, including 10,000 credits (about 1,000 ten-step runs). Extra credits are $0.01875 each, or 10,000 credits for $125 as a top-up. 25 concurrent browsers, 5 Android and 5 iOS virtual devices, 5 phone numbers, 30-day results, a shared Slack channel.
3. Enterprise: Custom, test-based pricing. Adds SAML SSO, SCIM, audit logs, 1-hour mobile sessions, an SLA, and a named contact.

Recommended for?

1. Teams that want an AI-assisted E2E suite and will review healed steps instead of trusting them.
2. Web apps whose flows fit hosted Chromium and a credit budget.
3. Signup and email paths, using the included inboxes, with SMS only on the paid plan.
4. Not for hardware-specific mobile bugs, payment-device labs, or teams that must run tests entirely inside their own VPC on a self-serve plan.

Conclusion

Momentic rents an AI end-to-end runner billed in credits. The free 2,000 credits and free editor runs are enough to learn whether the step style fits your UI. Pay-as-you-go at $125 a month is the first plan that does not hard-stop and that adds virtual mobile devices. Prefer normal steps over recovery, because recovery spends credits and can pass a broken product. Keep a real device lab if emulators are not the question you are asking.