Motion Studio

Write it. Watch it move.

Describe it. Opus 5.5 animates it in code, sound included.

Start from an example

How it works

1

The model designs it in code

Opus 5.5 writes the whole piece as one web page: layout, typography, illustration, Canvas or WebGL scenes, easing and transitions, plus a soundtrack built from oscillators and noise in Web Audio.

2

Every frame is rendered exactly

A headless browser steps the animation frame by frame at 30 fps and records the soundtrack offline, so nothing stutters however complex the scene, and the sound lands on the beat.

3

Watch it, then download

Open the live preview as soon as the code is written, scrub through it, and download the finished MP4 at 1080p in any of seven frame shapes.

FAQ

What is Motion Studio?

Motion Studio is CocoViral’s AI motion graphics generator. You describe a piece in plain words and Claude Opus 5.5 designs it as code: layout, typography, illustration, Canvas or WebGL scenes, easing and transitions, plus a soundtrack. Every frame is then rendered to a 1080p MP4. Fable 5.5 is coming soon as a second model.

How is this different from an AI video model?

Video models paint pixels, so text often comes out garbled and logos drift. Here the animation is written as code, so every word, number, logo and chart is exact, and the style can be anything from flat design to stained glass to 3D.

What can I make with it?

Kinetic typography, logo reveals, explainers and data animations, product launch teasers and end cards, title sequences, lower thirds and social promos. Anything that is mostly type, shapes, charts and illustration works especially well.

Which formats and lengths are supported?

16:9, 9:16, 1:1, 4:5, 4:3, 3:4, 21:9, from 5 to 60 seconds. You get a 1080p MP4 at 30 fps with the soundtrack mixed in.

How do I write a good brief?

Say what appears on screen and in what order, give the exact words in quotes, name a style or reference (Swiss poster, neon, paper cut-out) and the mood of the music. Short briefs work too: the model fills in the rest, so "go all out" is a valid instruction.

Does it come with sound?

Yes. The model composes music and sound effects in code and cuts them to the picture, so hits land on the transitions. Turn Soundtrack off for a silent video, or attach your own track. The live preview plays without the generated sound; the MP4 has the full mix.

Can I use my own images, video or audio?

Yes. Attach up to 3 images, 2 video clips and one audio track, or drop them anywhere on the page, and add a note to say what each one is. The AI looks at every attachment and builds the piece around it. Video is converted to WebM in your browser and its first 60 seconds are used. An audio track, such as a voiceover or a song, replaces the generated soundtrack: the animation is cut to its beats and words and runs no longer than it. Each attachment adds a little to the price: 2 credits per image, 5 per video clip and 15 for the audio track.

How long does it take?

Writing takes a few minutes, then rendering takes about as long again, depending on length. You can open the live preview as soon as the code is written, while the video renders.

What does it cost?

450 credits to write the piece plus 2 credits per second of video, so a 15-second video is 480 credits, plus a few credits for each attachment. The price is shown on the button before you start, and failed runs are refunded automatically.

Can I change a video after it is made?

Yes. Press the edit button on any finished video and it opens in the Agent. Tell it the change in chat, like "make the title bigger" or "swap the palette to blue", and the same model edits that video’s code and renders it again as a new video, keeping everything else, including your attachments. An edit costs 300 credits plus 2 credits per second, and you can keep chatting to edit the newest version.

Can I use it in the Agent?

Yes. Ask the Agent for motion graphics, kinetic type, a logo reveal or an explainer, and it writes and renders one for you. Ask it for changes afterwards and it edits the same video.