CineMorph · Video voice replacement

Re-Voice

Replace the voice. Keep the picture.

Convert the performance into a new voice, or rewrite the line and lip-sync the clip to it. Runs on your own GPU when the local engine is up; falls back to your own provider keys when it isn't.

Cloud · BYOK

1 · Clip

Drag & drop a clip — or

Choose clip

MP4 / MOV / WebM · one speaker · up to 60 seconds, 200 MB · best under 20 seconds

Re-Voice replaces the complete source audio track. Music, ambience, and effects in the clip are replaced too — keep them on separate tracks in your edit.

2 · Voice

Voice target

Reference sample Preset Describe a voice

Drop or choose a clean sample

3–60 seconds. Best: 3–15 s for Qwen, 10–30 s for VibeVoice.

Reference transcript (editable — improves clone quality)

3 · Mode

Re-voice — rewrite the line, new performance, lip-sync Convert — keep the performance, change the voice

Transcript (edit the words to change the line)

Advanced

Model

Language

4 · Render

MP4 — lip-synced video WAV only — replacement audio, no lip-sync