Start with a main voice

Paste your script and choose a language and main voice. TTS Lines splits the script into editable lines. Each line uses the main voice until you choose a different speaker for it.

Open the voice library to hear a sample of each voice. The speakers available on a line match the selected language. Choose Hebrew in the language menu to use one of the four Hebrew voices.

Change a speaker on one line

Use the Speaker menu below a line to assign another voice. Generate that line to hear the change. You can also change its speed or wording without starting the whole recording again.

Lines keep their speaker choices when you move them. Changing the main voice updates lines that still use Main voice; lines with their own speaker keep that choice.

Export the finished conversation

Play the full recording to check the handoffs between speakers. Download one MP3 or WAV for the complete script, or take individual MP3 clips into a video editor. The ZIP export groups the separate clips in line order.

The neural voice model runs in your browser after the initial download. Generation uses your device, so longer scripts and slower devices take more time.