---
title: "Detect human poses in an image"
method: POST
path: "/v1/pose/detect"
tags: ["Pose"]
---

# Detect human poses in an image

`POST /v1/pose/detect`

## Request

- Base URL: `http://localhost:9998`
- URL: `http://localhost:9998/v1/pose/detect`
- Auth: HTTP bearer

## Request body

- object
  - `image` string, required: Base64-encoded image
  - `model` string: Pose model id — an Apple Vision preset (`apple-body`, `apple-body-hands`, `apple-wholebody`, `apple-body-3d`) or a Core ML model id. Omit to use the active preset. This is the only detector selector.
  - `renderer` 'openpose' | 'overlay' | 'json' | 'depth'
  - `topology` 'openpose-18' | 'coco-17' | 'openpose-hands' | 'dwpose-133'
  - `include_hands` boolean: Detect hand keypoints. Omit to keep the selected model's setting.
  - `include_face` boolean: Detect face landmarks. Omit to keep the selected model's setting.
  - `detect_3d` boolean: Detect 3D body pose. Omit to keep the selected model's setting.
  - `confidence_threshold` number: Minimum keypoint confidence. Omit to keep the selected model's setting; a value outside 0–1 or a non-number is refused with 400.
  - `canvas_width` integer: Output canvas width in pixels, clamped to 64–8192. Omit to keep the selected model's setting.
  - `canvas_height` integer: Output canvas height in pixels, clamped to 64–8192. Omit to keep the selected model's setting.

## Response `200`

Detected poses

---

[API](https://skmtc.dev/modelpiper/apis/toolpiper-api.md) · [All operations](https://skmtc.dev/modelpiper/apis/toolpiper-api/llms.txt) · [OpenAPI document](https://skmtc.dev/modelpiper/apis/toolpiper-api/revisions/6722bf56385c?raw)