A focused sample of observable answers, competitors, and citation patterns. It is a one-time diagnostic, not recurring monitoring or a claim about every model.
Response-level answer counts across the available public-audit model set.
Full citation totals are retained even when only a smaller top-source list is displayed.
The run is durable, the link is shareable, and your submitted email is never returned by the public API.