How to Make AI Avatar Videos
An avatar on a blank background is just an avatar. Over your real footage, it's UGC.
AI avatars make it possible to have a consistent on-screen presenter without filming yourself. But a talking head floating on a studio backdrop reads as synthetic and rarely performs as social content.
This guide covers making AI avatar videos that actually work in a feed — by placing a lip-synced presenter over your own screen recording.
Step by step
- 1Write a short, spoken-style scriptAvatars sound best delivering conversational lines. Keep sentences short and front-load the point.
- 2Provide footage for the avatar to present overRecord your screen — the product, app, or workflow — so the avatar has something real to present, not a blank background.
- 3Generate the avatar and composite itMenivor creates a lip-synced HeyGen v2 avatar from your ElevenLabs voiceover and chroma-keys it over your recording in a picture-in-picture, hook-intro, or split layout.
- 4Add captions and export verticalKaraoke captions are burned in and the whole thing is composed into a native 1080×1920 MP4 in a single render.
Keeping the avatar in sync
The avatar's mouth, the voiceover, and the captions are all timed from the same ElevenLabs word-level timestamps, so nothing drifts. That shared timing is what makes the presenter feel convincing rather than uncanny.
Frequently asked questions
- Do I need a GPU or a HeyGen account?
- Not necessarily. Menivor can route the avatar step to a self-hosted lip-sync path when needed, billed through the wallet, so you don't have to manage infrastructure.
Tools for this
More guides
Put this guide into practice — make your first reel free
Start with 3 free reels. Success-only billing means you only pay when a render finishes, and credits never expire.