Two tools, one scorecard, verdict first. Every number in a comparison is traceable to the same fact base as the main ranking; cells still under verification are marked pending rather than filled.
A worked base-rate example for AI detectors: why a 90% sensitivity and 95% specificity still flag many honest writers, drawn from FavTutor's 2026 review and its TOEFL study figure.
Originality.ai pay-as-you-go against Pro: eesel AI's $30 credit pack, $14.95 Pro plan and credit rules, plus the 98% peer-reviewed result FavTutor reports for the tool.
Pangram vs ZeroGPT on what their pages state: Pangram's 2,000-word daily free plan and $20 Individual tier, ZeroGPT's 15,000-character no-signup check, and each vendor's own accuracy wording.
Six questions for a school before it relies on Turnitin's AI writing report, drawn from FavTutor's 2026 field review and eesel AI's June 2026 review of the false-positive record.
Winston AI vs GPTZero on independent evidence: FavTutor's 99% peer-reviewed result for Winston against eesel AI's 95.7% RAID figure for GPTZero, plus each vendor's own accuracy claims.