Live demo

Experience Vidu‑S

Watch it in action, try a ready-made character in one click, or bring your own photo and voice — then talk to it live.

The product demo video will play in this frame. Share the file or a link and it goes live on the next deploy.

Real-time generation — streamed, not pre-rendered · 540P / 25FPS · Two-way — it sees and hears the user

🎭 Ready-made character

Kai

Friendly, easy-going host

Speaks 29+ languages

你是凯(Kai),一个只会说中文的中国主播。请务必全程只用中文交流,不要说任何英文单词。你的第一句话必须是用中文说"你好呀,很高兴见到你!"来打招呼。之后回答要简短、自然、亲切,并对你看到和听到的内容做出回应。

Voice swap the voice, hear it first

Call mode

Checking demo availability…

IDLE

00:00

Hit Start call — a session spins up, the media channel opens, and Kai comes on-line, usually within seconds.

Runs against the real Vidu-S API through our secure proxy — the key stays server-side. Live calls bill 3 credits / 2s and cap at 600s in beta.

Photo one person · PNG/JPG · ≤18MB

Drop a photo or click to uploadFull or half body, any style — real, anime or pet

or paste an image URL

Name ≤ 20 chars

Persona who is this character?

Voice 56 built-in — or clone your own

Clone a custom voice

Voice name

Audio sample URL mp3 / wav

Language

First 10 clones are free, then 899 credits each. Point us at an audio link — direct file upload for cloning lands with storage.

Call mode

Add a photo to begin.

IDLE

00:00

Upload a photo, pick a voice, and hit Start call to meet your own digital human.

Your photo is sent inline (Base64) — nothing is stored. Calls run through the same secure proxy and bill 3 credits / 2s.