Listeners hear two AI voices read the same sentence and pick one, without ever seeing who made either. This is the tally. Filter by language and by what you actually care about — and look at the error bars before you trust a ranking.
Scores come from a standard head-to-head method called Bradley–Terry, centred so the average voice in each language sits at 1000. Error bars are worked out by resampling listeners rather than individual clicks, because one person voting 25 times is not 25 independent opinions. Results are frozen once voting closes; we do not publish a moving scoreboard while people are still voting.
Read the full report and method → · Download the raw results → · Get paid to be a listener →
