Script → Talking-Head Studio
Upload a script → per-speaker voice + talking-head clips → assembled film
⚙ Settings
1 · Speakers
headshot + voice sample per character
+ Add speaker
2 · Upload audio
skip the script — attach recordings, STT transcribes them
Speaker
Audio files (one line each)
Choose audio…
no files selected
Transcribe → lines
Upload pre-recorded audio, transcribe it, and generate heads directly — no script needed.
3 · Script
SPEAKER: line, [SPEAKER] line, or free text
or upload .txt
Choose file…
Parse → sequences
4 · Sequences
Assemble into one film
Generate
5 · Output
Nothing generated yet.
Film
⚙ Settings
fal.ai key
(key_id:key_secret)
👁
Test fal
Powers F5-TTS voice + HeyGen avatar4 heads. Render billing needs a funded balance (was previously exhausted).
LLM key
(script parsing)
👁
LLM base URL
LLM model
Fallback voice language
Test LLM
Save settings
Close