API OCR quality is not same ocr.space

I scan the same document in on ocr.space and in example C#. If I look do json result of ocr.space is quality much more better than example C# procject.
C#: RacycJirg.ppi.
ocr.space: Recycling-Papier (correct)

Same words I missing in C# sample too.

It’s normally???

You need to use SCALE=true. See here: Online test result varies from the API tests