Template faces everywhere
Stock avatars are available to everyone; the same face shows up in other people’s videos, so nobody remembers it as yours.
Custom Digital-Human Videos:One Photo and a Script Are Enough
Many organizations want a digital human to explain policies, products or history—then find the template avatars all look alike, the lips don’t match the words, the voice sounds robotic, and the face changes from scene to scene. Not good enough for anything official.
We build the character from scratch for you: a digital twin of a real staff member, a fictional presenter or a historical figure. You send photos or a brief plus the script; we generate it with the latest models, match voice and lip-sync, and deliver a subtitled film.
ForGovernment bodies and institutions that publish explainers year-roundCompanies that want a brand spokesperson characterMemorials, museums and cultural programsTeams going global that need multilingual presenters
01CHALLENGES
Where this kind of video usually gets stuck
Unnatural lips and voice
Mouth movements drift from the words and the voice sounds flat, which looks careless at a formal event.
The face shifts between scenes
The same presenter looks different in the studio, outdoors and in the hall—especially obvious across a series.
02WHY AI
How AI solves them
A character that’s yours
Face, hair, outfit and manner are designed for you—or generated from a real person’s photos with their consent—so no one else has it.
Voice and lip-sync tuned
We pick a voice by age, gender and tone, or use an authorized real voice, and align lips and pauses line by line.
Any setting you need
Your digital human can present from your lobby, workshop or scenic spot, rebuilt from photos—no studio set.
One character, a series
Once the character is set, every episode reuses it with the same face, clothes and voice.
03WHAT WE MAKE
What we can make
Policy and service explainers
A presenter walks through a new policy or procedure, with charts and on-screen steps.
Brand spokesperson and product intros
Your own spokesperson character introduces new products, the same face for every holiday and launch.
Historical figures who speak
Figures rebuilt from historical sources read their letters and tell their own stories.
Multilingual versions
The same content in Chinese and English, lip-synced to each language for overseas audiences.
04HOW IT WORKS
How it works
Each step comes to you for approval before we move on.
- STEP 01
You share the brief
Tell us who presents, to whom and where it plays, plus photos or references for the look and a script or key points.
- STEP 02
We set the look and voice
We show several character stills and voice samples; you choose one, and we keep refining until it’s right.
- STEP 03
We generate and sync
We generate the presenter speaking your script, add backgrounds, subtitles and graphics, and cut a first draft.
- STEP 04
Final delivery and more episodes
We deliver landscape and vertical cuts; for each new episode, just send the next script.
05CASES
Work we’ve done
Click through for the full film and project review.
Spaces & Events·Dec 2025
Red Letters: A Dialogue Across Time
Three martyrs step out of their frames to read letters and answer the host, on a screen about 29 meters wide.
See the full case
Public Interest & Public Communication
Southwest Zhejiang in Family Letters
An AI docent leads viewers through Liu Ying’s letter in a livestream style; both the docent and the historical figure are generated.
See the full case
Culture & Art·Dec 2025
Baidu Baike — Masterpieces Reimagined MV
Figures from paintings and artifacts rap while keeping their original look—the same approach we use for custom digital humans.
See the full case06DELIVERABLES
What you get, what you prepare
You’ll receive
- Digital-human film (landscape and vertical)
- Final character stills and design sheet
- Voiceover and subtitle files
- Chinese, English and other language versions
- Intro, outro and graphics templates
Before we start
- Where it will play: screens, website, WeChat, video channels or live events
- What to say: a few lines, a brochure or existing copy
- What you have: photos, logo, past videos and references
- Launch date and rough length
For process, pricing and revisions, see Process & Pricing.
07FAQ
Frequently asked questions
- What do I need to prepare?
- For a real person’s twin, a few clear front and half-body photos, with their consent; for a fictional character, a description or reference images. Plus a script or key points.
- Can the digital human speak English?
- Yes. Chinese and English are both supported, with lip-sync done separately for each; for other languages, message us on WeChat.
- How long does a film take?
- The first time we set the character takes a bit longer; a 1–2 minute film usually reaches first cut in about a week. Later episodes go faster. Rush production is available; timeline and price are agreed case by case.
- How is it priced?
- Message WeChat avx358 with the use, length and number of videos; we’ll reply with an approach and quote within 1–2 working days. Not sure yet? Start with a short sample.
Don’t see your need? Message us
Listed here is what we’ve done and what people ask most. For other industries or uses, add us on WeChat and send a one-line brief—we’ll reply with an approach and quote within 1–2 working days.
Page updated 2026-10-24