Unlike traditional motion design software, Jitter does not use keyframes but rather instructions to layers.
By instruction, we mean an operation that has one purpose to set an end value. As an example, you can take a look at the way Webflow, Protopie, or Keynote handle animations (as opposed to AE or Rive where you have to keyframe the properties of the layers).
Basic instructions — move, scale, rotate, change color, change opacity, etc. — are combined and repackaged to form bigger instructions (e.g. a Slide in is made of Move and Change opacity).
This method lets you hide the complexity of the animations without sacrificing creative control. Our AI Superagents work with the same operations, so that whatever they build can be handed off seamlessly to the designer for finishing touches. Nothing happens in a black box.
Common mix-ups
Expectation | In Jitter |
Trim a layer’s visible range on the timeline | Use In/Out (or opacity) animations instead |
Keyframe every property at a time | Add an animation that transitions from one state to another within a set duration |
Design changes only affect “frame 0” | Design is the resting state for the whole scene; animations overlay it |
Timeline rows are layers | Timeline rows are animations targeting layers |