Why this exists
Every "best AI website builder" article on the internet is written from the same three screenshots and the same affiliate dashboard. None of them tell you whether the thing actually holds a booking, survives a reload, or costs $40 in credits to finish. That is a measurable question. So we measure it.
How we work
- One spec per test. The same brief, versioned and public, for every builder.
- Binary criteria. A feature works or it does not. No 7.5 out of 10.
- Artifacts, not adjectives. Screenshots, logs, timings and billing lines sit next to the result.
- Date everything. Tools change weekly, so an old result is kept and dated rather than rewritten.
Independence
The site is funded by affiliate commissions. That is disclosed on every page, and it is the reason the methodology is public: if a commission ever shaped a result, the artifacts would show it. Nobody can buy a ranking position here, and no builder has review approval.
What we are not
Not a news site, not a course, not a community. One job: build the same thing in every builder and write down what happened.
Corrections
If something here is wrong, we re-run the test and log the change in thechangelog. Tell us at contact.