Talking-head clips with a spoken line
Most short-form content is a person talking to a lens for eight seconds. Text to Video makes that clip from a description: who they are, where they sit, what they say. Put the line in quotes and keep sound on, and the clip gets a voice for it. Use your AI influencer and the face is one you own.
Short lines land best — one or two sentences per clip. A longer script becomes a three-clip series with the same setting words, which cuts together as one video. The starter prompt below opens vertical at Standard with a line ready to edit.
Starter prompt
“A young man with locs and glasses sits at a desk with a microphone, looks at the camera and says "three things I wish I knew before I started"”
What this solves
- Tips and explainers on days you cannot be on camera
- Hooks that need a face, not a caption
- UGC deliverables in a setting you do not own
- The same presenter across a week of posts
The steps
Write the setting and the line in quotes
The engine reads quoted text as dialogue and generates the voice with the clip.
Text to VideoOr film your AI influencer instead
A creator built from your photos keeps the same face every clip.
AI Influencer VideosReframe the poster frame for the thumbnail
The 9:16 poster becomes a 16:9 thumbnail for the long-form version.
Reframe for PlatformQuestions
How long can the spoken line be?
One or two sentences per clip reads most naturally. For a 30-second script, make three or four clips with the same setting and cut them together.
Does the voice match the person?
The clip generates a voice that fits the presenter you described. Describe age and tone in the prompt for closer control; use reference audio for lip-sync to a recording.
Can I use my own face?
Build an AI influencer from 2–8 photos of yourself and use AI Influencer Videos — same face every clip, rendered on the People tier.
Talking-head clips with a spoken line
Open with these settings