Upload a script, cast characters to TTS voices or your own mic, then rehearse scene by scene.
Import / Export
Import a script or saved rehearsal, or export the current session in any format.
Import from URL
Import from file
Export
Rehearsal library
No saved rehearsals yet. Parse a script below to start.
Drag & drop a script here.txt · .md · .fountain · .pdf · .fdx · .osf · .reh
New script
Paste a script below. Supports standard screenplay format (INT./EXT. headings, CHARACTER name on own line, dialog follows) as well as the simple CHARACTER: text format. Lines starting with # are explicit stage directions.
Supports: • Screenplay: INT./EXT., CAPS names • Theatre: ACT I / SCENE 1 • CHAR: text · # direction • Import: .txt .fountain .pdf .fdx .osf
Cast your characters
Assign a TTS voice or your mic to each character. Add a speaking-style note per character so the voice matches their personality.
Match characters with voices— let an LLM cast every character automatically
1
Match from local library
The LLM picks the best voice you already have (cloned · designed · imported) for each character. No TTS engine required.
2
Match from online (fish.audio)
Search fish.audio's 2M public voices for the best fit per character and import them to your library. Needs internet.
3
Design new voices
Generate a unique voice per character from the LLM's description. Requires the Voice Design engine (Qwen3-TTS :8021).
0 / 0
0 / 0▶1■1
0 selectedClick lines to select · Shift-click range
The current TTS backend has weak style support — tone selections may have little or no effect.
Synthesizing
0 / 0
Characters
Input level-∞ dB
0:00
TRAINTurn 1 / 1
Cue
Your line
Read the lines above · press ▶ to hear the cue, then 🎤 to record0:00
Input level-∞ dB
Result
Script:
You said:
Session complete
All recorded and synthesized clips from this session.