Skip to main content
Momentic is an end-to-end testing platform for web, iOS, and Android whose AI agents author, run, and maintain the tests. The agents turn natural language into repeatable tests stored as readable YAML files in your codebase, and heal them when your UI changes. These tests are the verification layer for your application. Caching skips the AI call on steps whose locators still match. Run tests anywhere with the CLI: on your laptop, in a CI pipeline, or on a cloud agent sandbox. A team of agents built into the CLI can add coverage, debug failures, and maintain tests.

​
Run your first test

The wizard signs you in through the browser, installs the CLI, writes momentic.config.yaml and a sample test, installs Chromium, runs the sample test, and sets up every coding agent it detects:
To use a coding agent instead, paste the setup prompt into Claude Code, Cursor, Codex, Copilot, or any other agent. The agent runs the wizard with an API key and writes your first test. To test without test files or repo changes, Mo, the AI QA agent, tests a running application and reports the bugs it can reproduce. Sign up for the private beta.

​
What Momentic is good at

  • Regression coverage on every pull request. The same tests your coding agent writes run as CI checks.
  • Maintenance when the UI changes. AI test maintenance diagnoses a failing step, patches the test, and shows the diff.
  • QA with no test files and no repo changes. Mo tests a running app and reports the bugs it can reproduce.
  • One test language for web and mobile. Web tests run on Chromium, iOS tests on simulators, and Android tests on emulators. Steps, modules, and the CLI work the same way on all three.
  • The CI setups teams already run. Six recipes cover PR gates, preview deploys, nightly suites, SSO apps, third-party monitors, and tests generated from a diff.

​
For AI agents

  • llms.txt: an index of every page. llms-full.txt: every page in one file. Append .md to any page URL for clean Markdown.
  • skill.md: the capability summary for agents.
  • npx momentic skills --yes: install the Momentic skills into your repository (npx momentic-mobile skills --yes for mobile).
  • MCP server: a local stdio server that gives a coding agent a live browser or device plus tools to author, run, and triage tests.

​
Pick your platform

Web

Chromium-based browsers, locally or in CI

iOS

Local and remote simulators without real devices

Android

Local and remote emulators for APK-based apps
React Native, Expo, and Flutter apps use the iOS and Android quickstarts. Momentic tests the native build, and each quickstart shows the build command for your framework.

​
Learn the fundamentals

What is AI-native testing?

Use AI to author, run, maintain, and learn from tests

How Momentic works

The agent loop, auto-heal, and the step cache

Test structure

File format, steps, modules, and variables

Goal-based testing

Combine AI actions, precise steps, and durable assertions

​
Configure and run

Configuration

momentic.config, env vars, auth, AI

CLI reference

Every command, every flag

Results

Runs, videos, traces, and local reports

Common problems

Fixes for the errors support sees most

​
Reference repos

Examples

Working web, iOS, and Android projects, including GitHub Actions, CircleCI, and Bitrise workflows

Coding agent skills

Install the skill files with npx momentic skills so your coding agent knows how to write and run Momentic tests

​
Get help

Check Common problems for the errors support sees most often, with fixes. To reach a person, see Get help: email support@momentic.ai or post in your shared Slack channel. Working examples live in momentic-ai/examples.