How the 3 ranked AI training platforms score

Ordered by score. Pick a column heading to order by that figure.

#	Platform	Chance accepted	Expected pay	Worth an hour	Referral boost	Score
1	Anthropic	3.7%Nobody has told us how far they got there yet, so it is the average of every platform we can measure: 3.7%.	n/aThe board's pay could not be read.	n/aIts chance or its pay could not be read.	0%	n/aIts worth an hour could not be read, so it has no score.
2	Mistral AI	3.7%Nobody has told us how far they got there yet, so it is the average of every platform we can measure: 3.7%.	n/aThe board's pay could not be read.	n/aIts chance or its pay could not be read.	0%	n/aIts worth an hour could not be read, so it has no score.
3	xAI (SpaceXAI)	3.7%Nobody has told us how far they got there yet, so it is the average of every platform we can measure: 3.7%.	n/aThe board's pay could not be read.	n/aIts chance or its pay could not be read.	0%	n/aIts worth an hour could not be read, so it has no score.

The chance of being accepted is read, for a platform that pays us a referral fee, from the people its own referral list shows us, step by step from sign-up to accepted, and for any other from the Verified Experts who told us how far they got there: who was accepted, and who is still likely to be, a chance that falls the longer they have waited. Typical pay is read from the rates its board states, the same rates as every pay figure on the site, and a role that states no rate counts at the average. Each leans on the average of every platform in proportion to how little data of its own it has. The referral boost is the one part of a score with referral input: a platform that pays us no referral fee, or cannot take anyone new through our link right now, gets 0. How the ranking works.