Object Detection

Detect objects (e.g. welding defects) and return bounding boxes with confidence scores. Returns JSON with detections and optionally an annotated image.

POST/api/detect/
Request BuilderFill the form — code updates live
No file chosen
Image to analyse
Model id; defaults to best detector for your workspace
Add a token to fire a real call
curl -X POST "https://msightflow.ai/api/detect/" \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -F "file=@weld_sample.jpg" \
  -F "model=yolo_v12_weld" \
  -F "return_image=true"
ResponsePress Send to see live output
▶

No request fired yet.

Pick a file, fill the form, and press Send. Output appears here as JSON, an image (when the endpoint returns one), and raw response headers.