Video from words
Describe a scene. Chain the clips.
Pick a model, a shape and a length and press Make video. The video is made clip by clip, every clip starting on the last frame of the one before, so it can run for minutes without a visible seam, and the clips are stitched into one file. Start on a picture of your own, end on another, and remake only the clip you do not like.
Request access What makes the clips
Early access, and plain about it: today the default models are Previews simulated in your browser, not video models. The windows on this site show sample data.
Made bysimulated Previews todayVeo and Kling where a key is set
Before anything else
What is simulated.
We would rather say it here than let you find out in the app.
- Pre-betaThe Previews are the default makers today
Two simulated models paint clips in your browser from your prompt and a seed, with WebCodecs. They honour your start and end pictures exactly, because frame 0 is your picture. They are real video files and not video models, and they do not shoot what you described.
- Pre-betaTwo real models are connected in code
Google Veo 3.1 and Kling 2.1 work only where their keys are set on the server. Whether either is switched on for the hosted app is to be confirmed.
- Pre-betaCredits go further on the Preview
A second of a real model costs many times more credits than a second of the Preview, and the rates are not final.
What it does
Describe. Chain. Stitch.
The composer, the chain, the timeline and the library are built and in the app. They work the same with any model, and each clip says which one made it.
Make
A scene in words, or on a picture of your own. One clip, or a chain that fills a length.
Make videosChain
Each clip starts on the last frame of the one before.
The clip chainEdit
Reorder, duplicate, remake or cut clips. Only a changed clip is made again.
TimelineKeep
Downloads, pictures and Geenius OS Media with provenance.
Library
The chain
The last frame starts the next clip.
A model makes a few seconds at a time. To make a video longer than that, AI Vids makes a clip, takes its last frame and starts the next clip on it, again and again, and stitches the clips into one file. Fill a length in one go, or add clips one by one.
- Up to a minute a clip with the Long take Preview
- Real models make shorter clips and the chain does the rest
- A cut starts fresh instead of on the clip before
The timeline
Change one clip, make only that one.
Edit a clip's words, its length, the picture it starts or ends on. Reorder, duplicate or cut clips. A clip that changed is remade by the next make; the others are kept, so you do not pay for them twice.
- Remake a clip with a new seed
- Start or end any clip on a picture
- Download the video when it is ready
Who makes a clip
An open page makes it.
Clips are made by an open AI Vids page in your workspace: every page watches the queued videos and claims one, so two tabs never make it twice and a closed tab hands it on. Real models are called by that page through the server. So a video queued by someone else, or by your AI, waits until somebody has the app open.
- A lease of a few minutes, renewed while a clip is made
- A closed tab hands the video on
- No open page, no clips
Geenius OS Media
Saved with their provenance.
Save a video to your Geenius OS Media library with its story, every clip's prompt and length, the model and the settings, or bring a picture in from Media to start a clip on.
- The provenance travels with the video
- Pictures from Media or an upload
- Downloads any time
A door for your AI
Your AI queues. A page makes.
Connect Claude, ChatGPT or Grok to Geenius OS once. Your AI can ask which models are ready, start a draft video, edit its clips, queue the making and read the result.
- One tool, twelve actions
- On every plan, making on Premium and Max
- It cannot promise a result
In numbers
Facts you can check, not testimonials.
AI Vids is in early access and has no public customers yet. These are things you can look up in the app and the docs today.
- 4models: two Previews, Veo and Kling
- 2models ready for everyone: the simulated Previews
- 60 slongest clip, with the Long take Preview
- 12actions on the AI door
Honest labels
What works today.
We say what is live in early access, what depends on your setup, and what is not offered.
- Pre-betaComposer, chain, timeline, library and Media
Built and in the app now. Early access: people request access.
- Pre-betaYour AI through Geenius OS
Works when AI Vids is added to your Geenius OS workspace, on every plan.
- Pre-betaVeo and Kling
Connected in code and ready only where their keys are set. We do not claim they are on for you.
- Not offeredMaking videos on Pro
Pro is the library. Premium and Max make videos.
- Not offeredA phone app
Web only for now.
Who it is for
For people who want a chain and a library.
Animatics and mock-ups
Simulated clips for timing and drafts, never passed off as real.
See the setupLong chains
Chain clips on the last frame to fill a length and stitch them into one file.
See the setupGeenius OS users
Start clips on pictures from Geenius OS Media; save videos back with their provenance.
See the setupMakers with their own AI
Your AI asks what is ready, starts a video, edits clips and queues the making.
See the setup
Pricing
Three plans. No free tier.
Per month, in euros, VAT included. Included in the matching Geenius OS plan. Payments are in test mode during early access.
- €9Pro: the library, no making
- €19Premium: 1,000 credits a month
- €49Max: 3,000 credits a month
From the blog
Notes on video, honestly.
Why every clip starts on the last frame
How AI Vids makes a long video from short clips by starting each on the last frame of the one before, and what that does and does not give you.
Only remake what changed
How the AI Vids timeline keeps the clips you did not change, remakes only the one you edited, and why that saves credits.
Say what a preview is
Why AI Vids names its Previews as simulated, what they do, and why that is the honest starting point for video from words.
Questions
Good questions.
What is AI Vids?
A library and a maker for video. You describe a scene, pick a model, a shape and a length, and the video is made clip by clip, each clip starting on the last frame of the one before, then stitched into one file. Its full name is Geenius AI Vids. It is an app of Geenius OS and also works on its own.
Are the videos made by an AI model?
Not by default today. The default models are Previews that paint clips in your browser from your prompt and a seed; they are not video models. Google Veo 3.1 and Kling 2.1 are connected in code and work only where their keys are set on the server.
Will it make what I describe?
The Previews will not: they paint a scene that leans with your words. A real model can, where one is switched on.
Is it ready to use?
It is in early access: people request access and payments are in test mode. The composer, the chain, the timeline, the library and Media are built. Request access.
Who makes it?
AVS Media, a small company in Nijmegen, the Netherlands, as one of the apps of Geenius OS.
How does my AI use it?
Through Geenius OS: a tool called video with twelve actions to start a video, edit clips, queue the making and read the result.
Describe the first scene.
AI Vids lets people in step by step. Ask for a place, or sign in with Geenius OS.