The AI platform for human motion.

Generate editable 3D motion from video or text for games and simulations. Retarget it to supported characters, or work with Uthana on motion datasets and robot retargeting.

Developers: read the API docs or set up Uthana MCP

SOURCE VIDEO
STRUCTURED 3D MOTION
CHARACTER / ROBOT
For game studios and 3D platforms

Production-ready character animation.

Generate motion from video or text, retarget it to supported characters, and export editable FBX, GLB or BVH.

For simulation, robotics and AI teams

Human motion for simulation and AI training.

Populate simulations with animated humans, license 250+ hours of studio-captured motion, or scope a custom dataset or robot retargeting program with our team.

Built for physical AI and proven in gaming.

1,000+
hours of data to license across all collections
<50ms
speed to first motion token
45,000+
characters animated
1.7M+
seconds of motion generated
Unity AIReplicateUltimate BotsScenarioLayerSloydLudus AIBitMagicNilo
01 Models

Turn observation and intent into 3D motion.

Generate structured human movement from video, language, or real-time control.

3D motion from 2D video

Convert 2D video into editable, animation-ready or training-ready 3D motion.

Explore video-to-3D animation

Text-to-3D motion

Describe the action, sequence, pacing, posture, and style to generate editable 3D character motion.

Explore Text-to-Motion 3.0
“An archer takes two steps forward, then quickly draws a bow and fires it at an enemy, then takes cover”

Locomotion

Generate responsive movement from direction, speed and trajectory.

02 Developers

From prototype to production through one API.

Generate motion, extract 3D movement from video, retarget across bodies, and auto-rig characters — through one GraphQL API.

03 Data

Motion data ready for training.

License high-quality captured motion or turn real-world video into structured 3D movement data, annotations and model-ready datasets.

Studio motion dataset

Human movement captured in a AAA production studio using a 120-camera motion capture system.

Motion data from video

Transform existing 2D video into scalable 3D pose and motion data.

Labeling and enrichment

Add action labels, temporal segments, joint trajectories, contact states and custom metadata to match your training specification.

Delivery

Delivered in the formats your pipeline already reads.

Animation formats for production tools, tabular formats for training pipelines. Exact fields, skeleton conventions and units are documented per dataset before delivery.

FBX
GLB
BVH
0,1CSV
04 Embodiment

One motion. Many bodies.

Transfer human movement across different skeletons, characters and robot embodiments without rebuilding the motion by hand.

Motion retargeting

Transfer movement between different bipedal character skeletons.

Auto-rigging

Prepare unrigged 3D characters for animation automatically.

Robot retargeting

Convert human motion into robot-specific joint representations.

05 Solutions

Build systems that understand how bodies move.

Use Uthana to train robots, create interactive simulations, develop world models and power 3D experiences.

Interactive Simulation

Simulated humans that move with intent.

Build responsive human agents that act from goals, controls and environmental context — not libraries of pre-authored clips. Uthana is partnering with simulation teams to develop the next generation of interactive human agents.

Become a design partner

Robot training

Turn human motion into robot-ready data for simulation, imitation learning and control.

Explore robot training

World models

Add structured 3D movement and action labels to multimodal training data.

Explore world models

Game studios

Generate, retarget and ship production-ready character animation.

Explore Game studios

AI-native game platforms

Embed motion generation, rigging and retargeting directly into your product.

Explore AI-native game platforms
Pricing

Pay as you create.

Add credits when you need them. Each generation or operation draws from one balance across the web app and GraphQL API.

$0.10
per second of Text-to-motion 3.0 output
$0.06
per second of Video-to-motion 3.0 output
$0.02
per second of Locomotion output
$0.15
per auto-rigged character

No subscription required and no minimum commitment. Purchased credits do not expire. Datasets, data services, simulation, and custom integrations are priced to scope.

FAQ

Common questions.

How can I create motion with Uthana?

Use video-to-motion to reconstruct a performance from ordinary video, text-to-motion to generate an action from a written prompt, or locomotion to create controllable and stylized directional movement. Generated motion can be previewed in the browser, applied to a compatible character, and exported for use in a 3D workflow.

Do I need a rigged character to use Uthana?

No. You can test motion on a provided character, upload a compatible rigged character, or use auto-rigging to add a skeleton to an unrigged bipedal 3D character. Auto-rigging prepares the character for animation and retargeting.

Can Uthana apply motion to my character or robot?

Uthana retargets motion across different bipedal character skeletons and can adapt human motion to a target humanoid robot representation. Character retargeting addresses differences in bone names, hierarchy, proportions, and rest pose. Robot retargeting additionally requires the target joint structure, axes, degrees of freedom, and limits to be defined and validated.

Robot retargeting provides motion in the target representation; it does not by itself provide a complete control policy, dynamics model, or guarantee that the motion is executable on hardware.

Who is Uthana for?

Uthana serves robotics and AI labs, simulation teams, 3D platforms, game engines, studios, developers, and individual creators. Self-serve models and tools are available through the web app and GraphQL API. Datasets, custom data programs, simulation work, and enterprise integrations are scoped directly with Uthana.

Work with us directly

Co-develop custom datasets, embodiment systems, and simulation programs with our team.

Custom datasets

Capture, extraction, and labeling built to your training specification.

Integrations

Bring Uthana into your engine, pipeline, or platform through the GraphQL API and public plugins.

Embodiment programs

Retarget motion to your characters and robots, with support from our team.

Let's build to your needs

Get in touch