Preview. Every figure on this site is currently a fixture, not a measurement — the live crawl has not run yet. Nothing here describes Google's actual results.
Free key
method

Why we publish our own data instead of describing it

Every search API claims to be accurate and comprehensive. None of them shows you the results. So we show ours.

16 Jul 2026

Read any four search API homepages and you will find the same three adjectives: accurate, reliable, comprehensive. They are unfalsifiable, every vendor uses them, and nobody has ever chosen a provider because of them.

The alternative is arithmetic. 4,185 queries, 10 countries, 82,620 stored observations, 12 weeks, one endpoint — the same one on the pricing page. If the data is good, that is visible. If it is not, that is visible too, and we would rather find out in public than in a churn report.

What it costs us

Publishing the corpus means publishing the gaps. We cover two surfaces where competitors cover twenty. We bill failed calls where SerpApi does not. Both of those facts are on this site, on the comparison pages, in our own words, because a comparison table with no losses in it is an advertisement and gets read as one.

What it buys

Something to link to. A search API cannot be evaluated from a feature list, and a trial key only tells you the response parses. A public series tells you what the thing actually returns, over months, on queries somebody else chose. That is the evidence we would want before spending money, so it is the evidence we publish.

Take the data. Attribution is the only condition.

Run the same measurements yourself

Free tier, no card. Every figure above came from one endpoint on a weekly schedule.

Get a free API key
Reading

More notes

Get this measurement every week

One email each Monday covering new notes: what changed in the results, which domains moved, and where AI Overview switched on or off. Unsubscribe in one click.