OCR Extraction Accuracy character level/word level or document level

Hi All,

I am evaluating OCR.space API to integrate with on of our solution, and looking for high extraction accuracy.

Is there a way I can get the extraction accuracy from the API?
Character level/word level or document level.

There is no way to specific a generic OCR accuracy as it all depends on document quality.

On a high quality 300 dpi scan of a text book page the accuracy can be close to 100%, but on a 70 dpi scan of a distorted receipt it is much less.

=> The best solution is if you test your documents with our online ocr form. It uses - of course :wink: - our OCR API.