Turbo end-to-end testing
One click runs about an hour of all-out testing against your release candidate. You get a short report of what broke, with steps to repeat each bug.
It gets better over time
- Every bug becomes a test. When a run finds a bug, we keep a test for it. It runs on every later release, so the bug cannot come back quietly.
- New paths each run. The bots do not repeat themselves. Each run explores parts of your app that earlier runs did not reach.
- Focus on changed code. The run looks hardest at what changed in this release candidate.
What a run covers
- Random exploration bots. Bots click, type and navigate like impatient users, looking for crashes, dead ends and errors.
- Feature-flag combinations. We switch flags on and off in different mixes, because bugs hide where two flags meet.
- Old versus new. We compare your last release with the candidate and flag anything that changed in a way you did not intend.
- Browsers and sizes. Chrome, Firefox and WebKit, on desktop and mobile sizes.
- Languages. Missing translations, clipped text, right-to-left layouts, dates and currencies, time zones and daylight saving changes.
- Third-party chaos. We make Stripe, sign-in, email and maps slow, down or return junk. Your app must not lose data or charge anyone twice.
- Data and money safety. Double clicks, two tabs, retries, and making sure no user ever sees another user's data.
- Accessibility, speed and security. Accessibility checks, performance on slow phones and slow networks, and basic security probing.
- Real devices (premium). The same checks on real phones and tablets.
Pricing
Prepaid credits only. There is no free tier and no free credit.
| Run | Price |
|---|---|
| Turbo core | $200 per run |
| Add-on: languages | about +$100 |
| Add-on: chaos | about +$100 |
| Add-on: security | about +$100 |
| Add-on: real devices | about +$100 |
| Everything run (core and all add-ons) | about $500 |
What it is good at, and what it is not
Turbo testing is great at crashes, broken flows, layout problems and how your app behaves when things go wrong.
It cannot guess your business rules. If a discount should never go above 30%, or a refund must follow a certain policy, you need to write that down once. After that, we test it on every run.