Arena Text

In plain words · Which AI people preferred across every kind of question

What does it measure?

A popularity ranking of which AI's answers people preferred across every kind of question they actually asked. It is not an exam score but 'user preference', so it is the closest thing to real-world satisfaction.

Who evaluates it, and how?

On Arena Intelligence, a user enters a question, two anonymous AIs answer, and the user votes for the better one. Over 82 million votes are aggregated with a statistical model (Bradley-Terry) into an Elo-style score, and this table takes the 'style-controlled' version from the official dataset, which statistically removes the bias toward long, nicely formatted answers. This table is the combined score across all kinds of questions.

Reading the score

This is a head-to-head record, not an exam score (accuracy %). It is computed from the probability of being preferred over other AIs, so there is no maximum, and when a new AI enters, existing scores shift slightly. Typical values are 1,000–1,500; a 100-point gap means the leader is preferred about 64% of the time, a 200-point gap about 76%. Correctness is not counted — a wrong but friendly answer can win — so if accuracy matters, also look at accuracy tests (GPQA, AIME, etc.). Gaps of a few dozen points are effectively the same tier even if the rank differs.

Last updated: 2026-10-02