AI tools · weienwong.online
A prompt in. A finished video out.
Write what you want said, and get back a narrated, captioned MP4 sized for YouTube or Shorts — real footage under a real voice, cut to the script. Every model runs in this process on CPU: no GPU, no API keys, and nothing is sent to a third party.
One account works across every Weien Wong service.
What it does
A voice worth publishing
Kokoro for English narration that sounds like a person reading, Piper for about forty other languages. Audition any of them before you render.
Real footage, not a slideshow
Each scene gets moving stock video matched to what is being said — or a generated image, made on this machine, if you would rather nothing left it.
Edit the script first
The scene breakdown is yours to rewrite, reorder and trim before a single CPU second goes into rendering it.
Captions and a music bed
Word-level captions, a ducked music bed, and the whole mix normalised to the loudness YouTube expects.
Getting started
about a minute- Create an account. It opens right here — no redirect. The same account signs you in to every other service on the domain.
- Write the prompt. Paste the narration you want read, or give it a topic and let it draft the script.
- Pick a voice and a format. Preview the narrator, choose YouTube, Shorts or square, and set the captions.
- Render and download. Watch the progress per scene, then play the result in the page and download the MP4.
Before you publish
The narration is synthetic and the footage is licensed stock — check the credits the render produces and paste them into your video description where the licence asks for attribution. Anything you write is yours; make sure you have the right to publish it, and be straightforward with your audience about the voice being generated. Your projects and renders are stored against your account and are visible to nobody else.