OCR API

The OCR endpoint accepts a structured image annotation request body so mobile and backend integrations can migrate with minimal changes.

POST /api/vision/v1/images:annotate
Authorization: Bearer cb_live_xxxxxxxxx
Content-Type: application/json

{
  "requests": [
    {
      "image": {"content": "BASE64_IMAGE"},
      "features": [{"type": "TEXT_DETECTION"}],
      "imageContext": {"languageHints": ["eng"]}
    }
  ]
}