AI-native platforms · Uthana

The motion layer for AI-native gaming platforms

Uthana gives AI game creation and 3D creation platforms a motion layer they can call through an API. Send a generated character — rigged or unrigged — and get back animation on that character, retargeted and exported as FBX or GLB, ready to render in-product.

The gap

AI-native platforms can generate almost everything except motion

You no longer need to rely on a small set of canned animations or a basic controller for motion.

Generated characters still need motion

A character without animation is a static asset. Most platforms stop short of solving this.

Stock libraries don't scale with generation

A fixed set of walks and emotes doesn't match the variety of characters and worlds users are now creating.

Building motion infrastructure in-house is a major lift

Training motion models, building retargeting, handling export formats — this isn't a side project.

How it works

From generated character to animated character in one API call

Your platform generates a character and sends it to Uthana. Uthana rigs it (if needed), generates motion from a text prompt or video reference, retargets it to the character, and returns it as FBX or GLB.

Step 1 — Generate the character

Your platform creates or receives a 3D character from its existing asset pipeline.

Step 2 — Auto-rig if needed

If the character doesn't have a skeleton, Uthana rigs it automatically. Any bipedal humanoid mesh works.

Step 3 — Generate motion from text or video

Pass a text prompt or a video reference alongside the character. Uthana generates motion and retargets it to that character's proportions and skeleton.

Step 4 — Get back the animated character

Uthana returns the character with motion applied, exported as FBX or GLB. Ready to preview in your viewer or use downstream.

Why Uthana

What Uthana adds to your stack

An animation layer that matches the speed, depth, and integration model your platform already runs on.

Speed

Animation in seconds, not hours

Motion generation runs in seconds. Fast enough to sit inside a real-time product flow, not a separate offline step.

Auto-rigging

Handles unrigged characters

Most AI-generated characters come out unrigged. Uthana rigs any bipedal humanoid mesh automatically before animating.

Retargeting

Motion on the exact character

Not a generic skeleton with motion pasted on. Uthana retargets to the actual character your user created — proportions, skeleton, joint structure.

Depth

Beyond stock animation sets

Text-to-motion and video-to-motion generate a far wider range of behaviors than any fixed animation library.

API-first

Built for integration

Designed to be called programmatically from your product's backend. No separate UI, no manual workflow, no plugin dependency.

Completeness

Close the loop from asset to output

Your users go from character generation to animated, exportable output without leaving your product. No dead end at motion.

What this enables

What your users can build with motion

When animation becomes part of your generation stack, the output changes.

Playable characters

Users animate the characters they generate. Not static previews — characters that move.

Populated worlds

NPCs and background characters with varied motion, not identical idle loops.

Expressive avatars

Character creation that goes beyond appearance. Users get motion that matches the character they designed.

Get started

Add Uthana's motion layer to your platform

API access for auto-rigging, text-to-motion, video-to-motion, and retargeting. FBX and GLB output. 1M+ motion seconds generated. 15,000+ characters rigged and animated. Enterprise options for volume, latency, data isolation, and custom models.