A model's score can move because the test changed. Artificial Analysis' Intelligence Index v4.2 update added complex knowledge work and professional document reasoning, removed a saturated science benchmarkA standard public test set for comparing AI models — the shared scoreboards behind every "model X beats model Y" claim.Full definition →, and increased the weight of private test sets.
Each change alters which abilities contribute to the headline number. Private questions also reduce opportunities to tune directly to known tests, although they do not eliminate every form of gaming. The update is useful context windowThe maximum amount of text a model can consider at once — its working memory for the current conversation or task.Full definition → for interpreting a release comparison: compare models within the same index version before treating a score difference as capability progress. A stable model can receive a different number under a revised evalA repeatable test for AI quality — a set of tasks plus scoring — used the way software teams use test suites, because model output is too variable to judge by eyeballing.Full definition → mix.
Terms in this piece · Glossary
benchmark — A standard public test set for comparing AI models — the shared scoreboards behind every "model X beats model Y" claim.
context window — The maximum amount of text a model can consider at once — its working memory for the current conversation or task.
eval — A repeatable test for AI quality — a set of tasks plus scoring — used the way software teams use test suites, because model output is too variable to judge by eyeballing.
