Embodiment

Auto-rig bipedal 3D characters.

Upload a compatible humanoid character in FBX or GLB. Uthana estimates a skeleton and prepares the character to receive 3D motion.

A rigging model inside the motion pipeline

A static 3D mesh cannot receive skeletal animation until it has a compatible skeleton. Manual rigging requires joint placement, hierarchy setup, and skinning work before the character can move.

Uthana Auto-rigging estimates a humanoid skeleton from the uploaded character and prepares that character for motion. This is why the capability belongs in both Models and Embodiment: a model performs the inference, and the result makes the body usable in the motion pipeline.

Auto-rigging typically takes 30–60 seconds. The API returns an Auto-rigging confidence score so a platform can identify results that may need review.

Prepare a compatible character

The self-serve and API workflow accepts FBX and GLB character files smaller than 30 MB. Auto-rigging works best on recognizable bipedal humanoid characters with a visible head, torso, two arms, and two legs.

The model may have difficulty with:

  • Non-humanoid bodies, or missing major limbs
  • Extreme proportions that interfere with humanoid skeletal mapping
  • Intersecting or heavily occluded limbs
  • Long skirts, robes, capes, or other clothing that obscures leg separation
  • Large wings, tails, horns, or additional limbs
  • Poor finger geometry when finger joints are requested

Slightly stylized or cartoon characters may still rig successfully.

From mesh to motion

Upload the character and let Uthana create its skeleton. Then generate motion from text, video, or Locomotion controls; apply an existing motion through retargeting; preview the result; and download it for the production pipeline.

The rig works inside the same Uthana workflow as Video-to-Motion, Text-to-Motion, Locomotion, and motion retargeting.

Video-to-MotionText-to-MotionLocomotionMotion retargeting

Auto-rigging for 3D platforms

Character-generation, avatar, and UGC platforms can place Auto-rigging between mesh creation and animation. A static character enters the workflow; a character with a motion-compatible skeleton leaves it.

Use the GraphQL API to upload characters, enable Auto-rigging, request finger joints, and read the returned confidence score. Contact Uthana to discuss production volume, review thresholds, failure handling, and integration requirements.

API documentationTalk to us

Start with your character

Upload a compatible bipedal humanoid character, create its skeleton, and apply motion in the same workflow.

Start creatingAPI documentation