Sign Up

Avatar Talking Video

Create AI-powered talking avatar videos with your own script and voice.

This guide walks you through creating an AI-powered talking avatar video using Tornads. Unlike UGC Ads which generate everything from product info, the Avatar Talking Video workflow gives you full control over the script, avatar, and voice.

Overview

The Avatar Talking Video uses a progressive 4-step wizard. Each step unlocks after you complete the previous one:

  1. Choose Format — Select video aspect ratio
  2. Add Your Avatar — Upload your own image or choose from the gallery
  3. Add Your Script — Write a custom script or generate one with AI
  4. Choose Voice — Let AI pick a voice or upload your own voice sample

The estimated credit cost and video duration update in real-time as you configure each step.

Step 1: Choose Format

Select the video aspect ratio for your talking avatar video:

  • 9:16 — Vertical format, ideal for TikTok, Instagram Reels, and Stories
  • 16:9 — Landscape format, ideal for YouTube and website embeds
  • 1:1 — Square format, ideal for Instagram and Facebook posts

Once selected, a green checkmark appears and Step 2 unlocks below.

Step 2: Add Your Avatar

Choose who appears in your video. You have two options:

Upload Your Own

Click Upload Your Own to use your own photo or image as the avatar. Upload a clear, front-facing photo for best results. Supported formats include JPG, PNG, and WebP.

Click Choose from Gallery to browse the pre-built AI avatar collection. The gallery opens with a rich set of filters to help you find the perfect presenter:

  • Gender — All, Male, Female
  • Age Range — All, 18-30, 30-40, 40-50, 50-60, 60+ years
  • Ethnicity — All, White/European, Black/African, East Asian, South Asian, Hispanic/Latino, Middle Eastern/North African
  • Body Type — All, Slim, Average, Athletic, Muscular, Curvy, Plus Size, Petite, Tall
  • Style — All, Casual, Business, Formal, Sporty, Elegant

The gallery includes both a Public Gallery of pre-made avatars and a My Avatars tab showing any avatars you have previously saved.

Click on any avatar to select it. The avatar appears in the wizard with its image displayed.

Modify with AI

After selecting an avatar, you can click Modify with AI to alter the image using AI. Describe the changes you want (e.g., "change background to an office setting", "add sunglasses") and the AI will generate a modified version. This costs 5 credits per modification.

You can also click Change avatar to go back and pick a different one.

Step 3: Add Your Script

Write or generate the text that your avatar will speak in the video.

Custom Script

Click the Custom Script card to expand a text area where you can write your own script. The script must be between 10 and 500 characters. A character counter shows your current length (e.g., "219/500").

Generate with AI

Click Generate with AI to open a modal where you configure the AI script writer:

  • Prompt — Describe what the script should be about (5-500 characters)
  • Duration — Target video length: 15s, 30s, 45s, or 60s
  • Language — Select the script language
  • Target Audience (optional) — Describe who the video is for

The AI generates multiple script variants that appear as selectable cards. You can:

  • Browse through the generated options
  • Edit any script directly
  • Select your preferred script with a single click

Script generation costs 1 credit.

Script Language

Choose the language for text-to-speech. Available options:

English, Chinese, Spanish, French, German, Italian, Japanese, Korean, Portuguese, Russian

Caption Style

Select how captions appear in the video:

  • None — No on-screen text
  • Bold Impact — Large, bold text overlay
  • Minimal Clean — Subtle, clean text
  • Kinetic — Animated, moving text
  • Handwritten — Handwriting-style text

Step 4: Choose Voice

Select the voice that narrates your script. This step auto-completes with the default AI option.

Let AI Choose (default)

The AI picks the best voice to match your script and avatar. This is the recommended option and requires no additional configuration.

Upload Voice Sample

Click Upload Voice Sample to use a custom voice. Upload an MP3 file of at least 8 seconds (max 10MB). The AI will clone the voice characteristics for your video narration.

Generating the Video

Once all four steps show green checkmarks, the Generate Avatar Video button becomes active at the bottom of the page. It displays:

  • Estimated duration — Based on your script length (e.g., "~15s")
  • Credit cost — Varies by duration:
    • ~15 seconds — 35 credits
    • ~30 seconds — 60 credits
    • ~45 seconds — 85 credits
    • ~60 seconds — 100 credits

Click Generate Avatar Video to submit. Credits are deducted immediately and the video begins processing. You are redirected to the video detail page where you can monitor progress and watch the result once ready.

Tips for Best Results

  1. Use clear, front-facing avatar photos — The AI works best with well-lit, forward-looking images
  2. Keep scripts natural and conversational — Write as if the avatar is speaking directly to the viewer
  3. Match language to your audience — Select the language that matches your target market
  4. Try AI script generation first — The AI often produces engaging scripts; you can always edit them
  5. Choose caption styles that fit your brand — Captions boost accessibility and engagement
  6. Upload voice samples in quiet environments — For best voice cloning, use clean audio without background noise