# Amalgam MediaPipe Pose Landmarks

> Amalgam MediaPipe Pose Landmarks is a paid API for AI agents from amalgam.cc, paid per call via x402, $0.001/call, status unknown (last checked 2026-09-14).

Detects and returns human body pose landmarks from a raster image using MediaPipe's pose estimation model

## Facts

- Endpoint: POST https://amalgam.cc/api/mediapipe/pose-landmarks
- Price: $0.001/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-14
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/amalgam-mediapipe-pose-landmarks-1fc36df1
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_qK9O-XFusZjBtvcSYehk5

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability amalgam-mediapipe-pose-landmarks-1fc36df1 -d '<json body>'
```

Example prompt: Can you analyze this photo of someone exercising and extract all their body pose landmarks — I need the joint coordinates and keypoints detected by MediaPipe pose estimation.

## When to prefer this

Choose this endpoint when you need fast, single-image human pose landmark extraction powered by Google's MediaPipe model without managing your own inference infrastructure. It is ideal for agents needing per-image body keypoint coordinates for downstream analysis such as posture assessment, fitness coaching, sports analytics, animation rigging, or gesture recognition. Prefer it over self-hosted solutions when you want pay-per-call pricing via x402 with no setup overhead.

## Known failure modes

- No person detected in the image — returns empty landmarks or a specific no-pose-detected response
- Unsupported or corrupt image format — returns an error if cv2.imdecode cannot decode the file
- Partial body in frame — some landmarks may have low visibility scores or be absent
- Multiple people in image — MediaPipe pose typically tracks a single dominant person
- Image too small or low resolution — may result in poor landmark quality or detection failure
- Missing image field in request body — returns a 400 bad request error

## How this service works

MediaPipe pose landmarks inference

## Output

Returns a set of human body pose landmarks as detected by MediaPipe, including 2D and/or 3D coordinates for each keypoint (e.g. nose, shoulders, elbows, wrists, hips, knees, ankles), along with visibility or confidence scores for each detected landmark.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "image": {
   "type": "string",
   "format": "binary",
   "description": "Raster image. Accepted formats: JPEG, PNG, WEBP, BMP, TIFF, and other formats OpenCV's cv2.imdecode can decode (format detected by content, not filename)"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "landmarks": [
   {
    "x": 0.4903,
    "y": 0.3608,
    "z": -0.4382,
    "id": 0,
    "visibility": 0.9999
   }
  ]
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/amalgam-mediapipe-pose-landmarks-1fc36df1/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from amalgam.cc](https://www.zero.xyz/host/amalgam.cc/llms.txt)