What we know
2 reportedly scored 65 on the Artificial Analysis benchmark, which is claimed to be the highest score achieved so far according to two online sources. 2 scored 65 on Artificial Analysis. 2's score of 65 is the highest so far on Artificial Analysis.
Why it matters
2 is a language model developed by Anthropic. The Artificial Analysis benchmark appears to be a test or evaluation metric for AI models, but details about its methodology, credibility, and industry acceptance are not widely documented in mainstream or peer-reviewed sources. The claim of achieving the highest score so far is based on reports from secondary sources, specifically discussion forums and aggregator sites.
What is still unknown
Independent tests of accuracy, supported languages, and a broader release date are not in the reporting.