Output

A little curiosity. A new creation.

Give HY-Motion an input and see what it can do. Your result will appear here.

Meet the model

Create 3D character animation from a text description

Generate humanoid motion-capture animation from a text prompt (e.g. "walking", "sword attack", "dance") or a person video (model=gvhmr + videoUrl). Models: hy-motion (3 cr, default), kimodo-soma (3 cr), gvhmr (3 cr, video). Defaults: 3s duration, 60% intensity for text models. Returns motion JSON to retarget onto a rigged character (see model_rig). Async: returns a jobId — poll GET /api/v1/jobs/:id until it succeeds, then read result.animation.url.

What you can do

Create character motion using the selected motion model.
Inspect the model-specific inputs before choosing text-driven animation or motion reconstruction.

From experiment to application

Try a prompt in the playground, adjust the supported settings, then use the same model in your app. Your account balance works across the playground and API.

Connect an AI agent →

Creator

Tencent

Make your first request

Use your Sorceress API key on your server. The playground uses your signed-in account; your application uses the same model with an API key.

cURL request

curl -X POST 'https://sorceress.games/api/v1/tools/model_animate' \
  -H 'Authorization: Bearer $SORCERESS_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
  "model": "hy-motion"
}'

Full endpoint reference →

Inputs & settings

promptstring

Text description of the motion, e.g. "walking forward", "two-handed sword slash", "happy dance". Required for hy-motion / kimodo-soma; optional for gvhmr.

videoUrlstring

Public HTTPS URL of a monocular person video. Required when model is "gvhmr".

durationSecondsnumber

Clip length in seconds, 0.5–10. Default 3. Ignored for gvhmr (length follows the video).

intensitynumber

Movement intensity percentage, 10–100. Default 60. Higher = larger, more energetic motion. Ignored for gvhmr.

seednumber

Random seed for reproducible results. Default: random.

Full input schema

Input schema

{
  "required": [],
  "properties": {
    "prompt": {
      "type": "string",
      "title": "prompt",
      "description": "Text description of the motion, e.g. \"walking forward\", \"two-handed sword slash\", \"happy dance\". Required for hy-motion / kimodo-soma; optional for gvhmr."
    },
    "videoUrl": {
      "type": "string",
      "title": "video Url",
      "description": "Public HTTPS URL of a monocular person video. Required when model is \"gvhmr\".",
      "format": "uri"
    },
    "durationSeconds": {
      "type": "number",
      "title": "duration Seconds",
      "description": "Clip length in seconds, 0.5–10. Default 3. Ignored for gvhmr (length follows the video)."
    },
    "intensity": {
      "type": "number",
      "title": "intensity",
      "description": "Movement intensity percentage, 10–100. Default 60. Higher = larger, more energetic motion. Ignored for gvhmr."
    },
    "seed": {
      "type": "number",
      "title": "seed",
      "description": "Random seed for reproducible results. Default: random."
    }
  }
}

Results, saved runs & billing

Media requests return a job ID. Poll GET /api/v1/jobs/:id until it succeeds or fails; keep the same ID while it is processing. Long chat requests can also return a saved job to poll. Completed outputs are stored in your account, and successful work is billed using your model's current rates.

A multi-step workflow may include separately billed successful actions. If a connection drops, check your saved run before submitting it again.

Know the cost

Pricing that makes sense

Dollar values are the USD equivalent of your Sorceress credits.

100 credits = $1.001 credit = $0.01 USD

Sorceress credits

$0.03 / 3 credits

Per successful animation

1 credit = $0.01 · Successful work is billed on completion.

Review the supported inputs for duration, quantity and other usage-based charges before submitting. Read the billing guide →

HY-Motion API

Input