What is Trace OCR?
Trace OCR is a REST API for extracting text from images and PDFs. It combines deep learning-based text detection with text recognition to deliver structured OCR results with confidence scores and bounding box geometry. Upload a document and get back every word, its location, and how confident the model is — all in a single API call.What is OCR?
Learn what OCR is and see a real example of Trace OCR in action.
Text detection
Localize text regions and get bounding box coordinates without reading the text.
Text recognition
Read text content from pre-cropped images. Returns the recognized string and confidence.
Key information extraction
Extract structured key-value pairs from documents like receipts and invoices.
Quick example
Get started
Make your first OCR request in minutes.