Real presenters work the same way
Filmed footage takes the same captions, name tags and screen inserts: here a presenter on one half and the live website build on the other.
See the split-screen explainerA portrait that speaks.Your script, your voice, your format.
Give JsonCut a portrait and a voice track. Fabric 1.0 or Aurora turns it into a speaking presenter, up to 60 seconds, ready for captions, graphics and music on the same timeline.
- 60 sper take, at 480p or 720p
- 2models: Fabric 1.0 for lip-sync, Aurora with gestures
- 21voices to speak with, in 15 languages
Two models,two kinds of presenter.
Real sample frames from each model, as Studio shows them in the model picker.
Fabric 1.0VEEDA portrait speaks your voice track with natural lip-sync
AuroraCreatifyExpressive talking heads with gestures
What you can do.Without leaving the project.
Start from a portrait
A photo you own, or one you generate with any of the twelve image models in the same project.
Write or upload the voice
Type a script and pick one of 21 voices, or use a recording. The take follows the voice track word for word.
Finish it like any clip
Word-timed captions, a name tag, b-roll and music go on top as editable layers. Change one word without generating again.
Where it is used.Real projects, open them in Studio.
One portrait, another language
Mara again, now speaking German: the same photo, a German voice in the same tone, lips in sync and German captions timed to every word.
See voiceover and translationBefore you start.The questions we hear most.
Can I use my own face?
Yes, any portrait you have the rights to. The model animates the picture; it does not clone a voice. Pick a Studio voice or upload your own recording.
How long can a take be?
Up to 60 seconds per take, at 480p or 720p. Longer videos are several takes on the timeline, cut between b-roll and graphics.
Can I try it for free?
Building and previewing in Studio is free. AI generation, the AI Director and exports are part of the 14-day trial and the paid plans; credits show the cost before anything runs.