Eight voices invented from text descriptions alone — no recorded samples, no cloning. Same sentence, eight different speakers. Press play and compare.