Category measurement

Which payments apis companies do AI engines name?

Not a ranking of which company is better. A count of which companies ChatGPT, Perplexity, Claude and Gemini actually name when a buyer asks them to recommend a provider — across 115 measurements.

Provisional — built from 1 scan. A category’s shape needs at least three before these counts are stable.

What these two numbers are

Times named is what we observed about every company here: how often each was named across every measurement in this category. It is a count, not a score.

Cited Score appears only where a company has been measured in its own right — a full scan of the category’s query pack, three runs per query per engine, with the coverage gate applied. A company observed only inside someone else’s scan has no score here, and giving it one would be inventing a measurement that was never taken.

The full method, including the query pack, the model versions and the change log, is at methodology.

Measured something different, or think a company is missing? Tell us. Every figure traces to the specific answers that produced it.