Why We’re Building an AI-Native Motion Editor

Motion design has a strange problem.

The tools are incredibly powerful, but the workflow hasn’t changed much in decades.

You still create layers. You move them around a timeline. You add keyframes. You adjust easing. You repeat.

And when you want to make a small change across a complex animation, you often have to manually find the right properties and change them one by one.

At Premation, we think there is a better way.

Motion design should be programmable

Traditional video editors are built around the idea that humans manually manipulate every part of a composition.

AI changes that assumption.

Instead of thinking:

“How do I animate this?”

We can start thinking:

“What should this animation do?”

Imagine selecting a title and saying:

“Make this enter from the bottom with a soft spring animation.”

Or:

“Make these cards appear one after another with a 120ms delay.”

Or:

“Give this scene a more cinematic, slow-paced motion style.”

The goal isn't to replace the motion designer.

The goal is to remove the repetitive work between an idea and its implementation.

What is Premation?

Premation is an open-source, AI-native motion editor designed for modern motion design workflows.

It combines a familiar timeline-based editor with programmable animation systems and an AI assistant.

The editor includes:

• Timeline-based editing
• Layers and scenes
• Keyframe animation
• Motion presets
• Procedural animations
• Visual effects
• An AI assistant
• Extensible plugins
• Browser-based rendering
• Desktop workflows

If you've used tools like After Effects, the basic concepts should feel familiar.

But the underlying idea is different.

AI isn't an extra feature sitting next to the editor. It's part of the editing workflow.

From keyframes to intent

Keyframes are powerful because they give you precise control.

But they also create a huge amount of manual work.

For example, a simple animation might require changing:

• Position
• Scale
• Rotation
• Opacity
• Timing
• Easing
• Delays
• Staggering

A designer might spend several minutes building an animation that can be described in one sentence.

That's where an AI-native editor becomes interesting.

Instead of treating the timeline as the only way to communicate with the animation system, Premation can use natural language as another interface.

You can still edit every keyframe manually.

But you can also describe the behavior you want.

Precision when you need it. Intent when you want it.

AI should understand the animation, not just the prompt

One of the biggest challenges with AI-generated motion is consistency.

Generating one cool animation is relatively easy.

Generating a complete project where everything follows the same visual language is much harder.

That's why we're interested in building systems around:

Design tokens

Motion should have a visual language.

Things like:

• Duration
• Easing
• Spacing
• Scale
• Typography
• Colors
• Transition styles

can become reusable design rules.

Motion exemplars

Instead of asking an AI model to invent everything from scratch, we can give it examples of animations we like.

For example:

Example 1: Fast UI entrance Example 2: Soft spring interaction Example 3: Cinematic title reveal Example 4: Minimal product transition

The AI can learn the patterns behind these examples and apply them to new compositions.

This is much closer to how designers actually work.

We don't reinvent our visual language every time we create something.

Why open source?

Premation is open source because we believe creative tools should be hackable.

Motion design is becoming increasingly connected to code, generative graphics, WebGL, WebGPU, AI and interactive experiences.

A closed editor can provide a fixed set of features.

An open editor can become a platform.

Developers can build plugins.

Designers can create motion systems.

Researchers can experiment with new interfaces.

AI developers can build new workflows.

And users can inspect how their tools actually work.

That is the direction we want Premation to take.

We're still early

Premation isn't trying to pretend that we've solved motion design.

We haven't.

There are still difficult problems around:

• Complex compositions
• Advanced expressions
• 3D workflows
• AI reliability
• Large project performance
• Rendering
• Animation interoperability
• Professional production workflows

But that's exactly why we're building it in public.

The current version is an experiment in what an AI-native motion editor could become.

And we're learning from every person who downloads it, opens the editor, breaks something, creates something interesting, or tells us what doesn't work.

The future of motion design might be conversational

Imagine opening a blank project and saying:

“Create a 10-second product launch animation for a dark SaaS website.”

The editor creates the initial composition.

You say:

“Make the typography more minimal.”

It changes the typography.

“Slow down the opening.”

The timing changes.

“Make the cards enter with a subtle spring.”

The animation updates.

Then you take control.

You adjust the keyframes.

You change the easing.

You move things around.

You override the AI.

That's the important part.

AI generates possibilities. The designer remains in control.

We don't think the future is about pressing a button and receiving a finished video.

We think it's about creating a new interface between human intent and motion.

That's what we're building with Premation.

And we're just getting started.

---

Premation is open source and available at premation.com.

If you're a motion designer, developer, creative coder, or simply curious about where AI-native creative tools are going, we'd love for you to try it and tell us what you think.