AI UGC video
AI UGC video: a presenter who talks to camera
UGC-style video is one person talking straight to the camera, as if filmed on a phone. Root makes it two ways: a video model speaks your line itself, or you record the line and lip sync it onto a clip. A 15-second vertical talking clip comes to $3.00, presenter included.
Updated
First month free on the Beginner plan, with 650 credits to try any model.
AI video with sound, made in Root
One 15-second talking clip, two ways
Both start from a vertical 9:16 still of the presenter, made in Root. Prices are what Root charges, if each run works first time. 100 credits is $1.
| Way | Step | Model | Price |
|---|---|---|---|
| The model speaks your line | The presenter, a 9:16 still | Seedream 5.0 Pro | $0.06 |
| The talking take, 15 s, HD, with sound | FLUX 3 Video | $2.94 | |
| This way in all | $3.00 | ||
| Your own voice, lip synced | The presenter, a 9:16 still | Seedream 5.0 Pro | $0.06 |
| A take of them talking, 15 s, 720p, no sound | Kling 3.0 Omni | $1.45 | |
| Lip sync to your recording, 15 s | Sync 3 | $2.30 | |
| This way in all | $3.81 |
Which way to choose
- Let the model speak when you want it quick and the exact voice does not matter. The four Seedance models, FLUX 3 Video and MiniMax H3 (Hailuo 3.0) make speech with the picture.
- Use your own voice when it has to be your voice, your brand's way of saying it, or a line that must be word for word. Lip sync runs in the Root app.
- For lip sync, start from a take of the presenter talking, so the head and face move with the words. The AI lip sync guide has the other tips.
- Make the presenter still vertical from the start: on the four Seedance models and MiniMax H3 (Hailuo 3.0), the take follows the still's shape.
How to make a talking clip in Root
- Make the presenter. Add an Image Gen node, pick 9:16 and describe a real-looking person in a real place: “a woman in her thirties in a small kitchen, holding the phone at arm's length, daylight from a window”.
- Write one short line. Put the words in quotes and say how they are said: she smiles and says, “I didn't think it would fit in my bag, but it does.” One speaker, one or two sentences.
- Run the talking take. Add a Video Gen node with a model that speaks, wire the presenter in as the first frame, turn sound on and run it. The price shows first.
- Listen, then keep or redo. Check every word. If a word comes out wrong, run it again, or record the line yourself and lip sync it onto the take.
Related guides
UGC video questions
What is AI UGC video?
Video in the style of a customer's own clip: one person talking to the camera, as if filmed on a phone, usually vertical. With AI, the presenter, the room and often the voice are made by the model. On Root, a 15-second vertical talking take with FLUX 3 Video is $2.94.
Which model should I use for a talking clip?
For a vertical take you can post as it comes, FLUX 3 Video, which makes the voice in the same run. The AI video with sound guide lists the other models that speak.
Can I use my own voice in an AI UGC video?
Yes. Record the line yourself, make a clip of the presenter talking, and lip sync your recording onto it in a Lipsync node. Lip syncing 15 seconds with Sync 3 is $2.30. Lip sync runs in the Root app.
Can I keep the same presenter in every video?
Yes. Make the presenter once as a still, then start every talking take from that still. For a presenter who comes back often, the consistent character guide covers reference images and training.
How long can one AI talking take be?
Up to 30 seconds with Seedance 2.5, the longest of the models that speak. Keep each take to one or two sentences, though: a short take is easier to get right, and you can make several.
Can I make talking clips by API or from ChatGPT?
Yes, with the four Seedance models and FLUX 3 Video: they run in the Root app, by API, and from Claude and ChatGPT. MiniMax H3 (Hailuo 3.0) runs in the Root app only. Lip sync runs in the Root app, so the second way is finished there.
Can I try AI UGC video for free?
For a month, yes. The first month of Root's Beginner plan is free and comes with 650 credits, enough for two 15-second vertical FLUX 3 Video talking takes. After that the plan is $9 a month, and you can cancel whenever you like.
Try Root free for a month
The first month of the Beginner plan is free and comes with 650 credits to spend on any model. After that it's $9 a month, and you can cancel whenever you like.