Eleven v4 Turbo
Eleven v4 reads a script the way a voice actor would. It knows who's speaking, what just happened, and how every line should land.
Drop a script the app generated, or sign in and pick one from the scripts table.
Every generation starts with a voice. Pick the role you read; Eleven v4 Turbo reads the rest.
At ~150 ms median time to first speech, the latency disappears and the scene flows. Each partner line shows the time it took here.