Skylark Models

Quickstart

From nothing to a detection in about a minute.

1. Get a key

Create one on the API Keys page. Keys are shown once, so copy it straight away.

2. Pick a model

List them from the catalog, or browse them in the playground.

curl https://skylarklabs.ai/api/models/catalog | jq '.solutions[].id'

3. Run it

curl -X POST https://skylarklabs.ai/api/models/infer/image \
  -H "Authorization: Bearer $SKYLARK_API_KEY" \
  -F "solution=cctv_outdoor_vehicle_detection" \
  -F "file=@street.jpg"

Boxes come back both in pixels (bboxPx, in the resolution the service analysed) and normalised to 0–1 (bbox). Draw with the normalised pair so your overlay does not depend on the service's resize decisions.

4. Or from code

Every client is a single dependency-free file, served by the API itself, so the one you download always matches the build you are calling.

curl -O https://skylarklabs.ai/api/models/sdk/python
python3 skylark_models.py --help

On this page