German OCR — Extract German Text from Images

This page opens the image-to-text tool with German preselected — drop a scan or photo and press the button. The German model correctly recognizes the umlauts ä, ö, ü and the sharp s (ß), which plain English OCR replaces with a, o, u or B and quietly corrupts the text.

German is famous for long compound words like Grundstücksverkehrsgenehmigungszuständigkeit — the model handles them well because it reads letter shapes rather than dictionary entries, but sharp input matters: at low resolution the many tall letters of a compound start merging. Scans at 300 DPI give the best results for German documents, invoices and letters.

Everything runs locally: your document is never uploaded, only the German language model is fetched once and cached by the browser. Copy the result or save it as .txt when recognition completes.