Skip to main content
Your AI Twin is a digital version of you. It looks like you, sounds like you, and can deliver any script you give it. Record it once, and you can generate videos of yourself at any time, on any device, without ever turning on a camera again.

​
Step 1: Create your AI Twin

1

Open AI Creator

In the Captions app on iOS, tap AI Creator from the home screen.
2

Select Create an AI Twin

Swipe the carousel to Create AI Twin and tap it. Tap Get started.
3

Take your photo

Captions will ask you to take a photo of yourself smiling with your teeth showing. This is how the AI models your face.Tips for a good training photo:
  • Use good, even lighting (natural light from a window facing you works well)
  • Look directly at the camera
  • Smile naturally with teeth showing
  • No sunglasses, hat, or anything obscuring your face
  • Neutral background is easiest for the model to work with
4

Choose your voice

You have two options:Select a preset voice: choose from a library of voice options. Pick one that roughly matches your natural tone and cadence.Create a voice clone: record yourself reading a short passage. This captures your actual voice and produces a significantly more authentic result. Strongly recommended if you want the Twin to sound like you.To create a voice clone:
  1. Choose Clone my voice
  2. Read the provided passage clearly and naturally. Don’t perform, just speak normally
  3. The recording is used to model your voice characteristics
5

Choose your AI Twin variant

Captions generates several variants of your Twin. Review each one and select the one that looks most natural and accurate.
6

Name your Twin and save

Give your AI Twin a name. It will appear in your avatar list across all devices on the same account.

​
Step 2: Generate a video with your AI Twin

Once your AI Twin is created, it’s available anywhere you use AI Creator.
1

Open AI Creator

Tap AI Creator from the Captions home screen.
2

Write or record your script

Enter your script or generate one from a prompt.→ Script to Video: The No-Camera Workflow
3

Select your AI Twin as the avatar

In the avatar selection step, your AI Twin will appear alongside the community avatars. Tap it to select it.
4

Generate and review

Tap Generate. Your AI Twin will deliver the script. Review the output. If the delivery feels off, edit the script and regenerate.

​
Use cases

  • Consistent social content: generate daily or weekly videos without recording sessions
  • Course content: film an entire course once (creating your Twin) and update or expand it indefinitely
  • Multilingual content: generate the same video in multiple languages with your face on all of them
  • Ad variations: test different scripts and hooks without scheduling a shoot
  • Internal communications: send team updates as video without setting up a camera every time

​
Tips

  • Voice clone makes a big difference. The preset voices are fine for testing, but a cloned voice sounds unmistakably more like you. Set it up once and you’ll never think about it again.
  • Lighting and background in the training photo matter. A clean, well-lit training photo produces a more accurate and convincing Twin. It’s worth retaking the photo if the first attempt looks off.
  • Create multiple Twins for different contexts. If you have Max or Scale, consider creating a more formal-looking Twin for professional content and a casual one for social content.

Script to Video: The No-Camera Workflow

Full AI Creator workflow using your Twin.

Create Multiple Ad Variations Fast

Use your AI Twin to produce ad variations at scale.

AI Twin reference

Full feature documentation.

AI Creator reference

Full AI Creator documentation.
Last modified on April 20, 2026