Examples you can inspect.
Real, dated API observations with the input, protocol, query strings, source domains, token use and limits kept together.
How to compare OpenAI and Gemini fanout locally
02How can you compare OpenAI and Gemini for “best SEO tools” without inventing missing evidence?
How much does country change fanout queries?
03Does the country setting change the queries for the same keyword?
Why does the same keyword produce different queries?
04Can the same setup return different fanout queries?
Which sources appear for AI comparison questions?
What does an AI web-search run look for when comparing Ahrefs and Semrush?