Product
Blender
Motion BuilderDocs →
3ds MaxSoonDocs →
Models
KimodoArdy
Company
PricingDocsNewsCommunityAbout
Start freeSign in

Kimodo. Motion that respects your keys.

Runs inside

Kimodo is a text-conditioned diffusion model for 3D character motion, from NVIDIA's Toronto AI Lab. Give it a humanoid skeleton and a description, and it writes the motion as editable keyframes.

Animatica runs Kimodo behind the open MMCP protocol, so one model answers every Animatica plugin. Poses you set are hard constraints: the model fills the frames between them and never moves them.

Run it hosted on our GPUs, or self-host the Apache-2.0 reference backbone on your own machine. Same plugin, same protocol.

Trained by
NVIDIA Toronto AI Lab. Upstream research
Type
Text-conditioned diffusion model for humanoid motion
Output
30-fps motion as editable keyframes on your rig
Runs in
Blender, Motion Builder
Self-hosting
motionmcp-kimodo, Apache-2.0, over MMCP
Training data
Outside Animatica's control. See the model index.

What it does.

  1. 01

    Text-to-motion

    Describe the action and get clean 30-fps motion baked to your armature.

  2. 02

    Keyframe constraints

    Pin the poses that matter. Kimodo generates the frames between them and leaves your keys where they are.

  3. 03

    Path constraints

    Sketch a curve. The root locks to it and the body stays alive on top.

Try Kimodo in your scene.

Start freeView source