How to extract text from an image or scanned PDF
Choose OCR for pixels that contain text. Choose the right language and verify the words against the original.
Choose the input that fits the job
Use Image to text for a JPG, PNG or WebP. Use Scanned PDF to text when the source is a PDF; it renders your selected pages and recognizes their pixels. This does not preserve a table as a spreadsheet or make a searchable PDF.
Prepare a readable source
Start with a clear, upright image. The image tool accepts up to 8 MB and 6 megapixels. The PDF tool accepts up to 10 MB, 50 source pages and five selected pages per run. A large file may still fail on a device with limited memory.
Select the language and layout
Choose English, Persian or both based on the source. Automatic, single-block and sparse-text layouts are available. The sample is in English and selects English so you can compare a known input with the result.
Run locally and verify important details
Tesseract and its language files download from this site when you run OCR. The selected file is processed on your device. Check names, numbers, punctuation and reading order manually; an engine confidence value is not a guarantee of correctness. Copy the text or download TXT/JSON after checking it.