Fume Overview

Docs

Fume Overview

What Fume is, how it works, and why teams use it for effortless end‑to‑end browser testing.

Fume is your AI QA team. Describe what you want to test with a Loom video and Fume will create and maintain Playwright end‑to‑end browser tests for you.

With Fume you:

Record what success looks like once, then let Fume codify it as Playwright tests
Keep tests healthy with self‑healing updates as your UI evolves
Run tests in the cloud or your own CI, and own the Playwright code
Catch regressions earlier with realistic, runtime execution against your app

Why Fume

Eliminate manual test maintenance: Fume updates selectors and flows automatically
Author faster: English + product video in, robust E2E tests out
Run anywhere: use Fume's cloud runners (free) or your own CI machines
Migrate easily: bring existing Playwright/Selenium/Cypress tests; Fume will convert and maintain them

How it works (at a glance)

Record a short Loom showing the flow and expected outcome.
Fume analyzes the video and your app to generate Playwright tests.
Run the tests in Fume or your CI. On failures, Fume can fall back to a browser‑use agent to complete the run.
Fume updates the test code to keep it stable for future runs.

Add a simple diagram of the lifecycle with arrows: Record → Generate → Run → Self‑heal.

What you own

You always own the output Playwright code—even if you cancel your subscription.

Next steps

Quickstart: get your first test running in minutes
Concepts: learn the moving pieces (recordings, generation, runners, self‑healing)
CI/CD: run tests in GitHub Actions/GitLab/your CI
Migration: bring existing tests into Fume
Security: how we protect data

Quickstart