Read tutorials
Follow detailed PDF, invoice, table, and scan-quality workflows.
File support, Google Docs and Sheets output, scan accuracy, privacy, limits, and troubleshooting.
The website accepts PDF, JPG, JPEG, PNG, BMP, TIFF, HEIC, and WebP files. Use a clear source file and keep website uploads under 20 MB. ZIP files are not accepted by the web dashboard.
Yes. Choose document text output to recognize image-only PDF pages. See the scanned PDF to Google Docs tutorial.
Yes. Sheets output analyzes visual table structure and returns recognized rows and columns. It works best for clear invoices, statements, schedules, forms, and price lists.
Docs output prioritizes readable page text and paragraph flow. Sheets output prioritizes cells, rows, and columns. Choose the format that matches how you will use the extracted information.
Use a straight, focused, high-contrast scan around 300 DPI. Remove shadows, rotate pages, crop unrelated backgrounds, and use the original PDF instead of a screenshot when possible. Read the OCR accuracy guide.
No OCR system is perfect. Always review names, dates, account numbers, invoice numbers, totals, medication details, legal clauses, and other important fields against the original document.
Blur, skew, wrapped labels, merged headers, faint spacing, and angled photos can make boundaries ambiguous. Crop the table, straighten the page, and process the clearest available source.
Handwriting recognition varies significantly with writing style, spacing, image quality, and language. Neat block writing usually performs better than connected cursive. Review handwritten results carefully.
No. Website upload storage is private. The service uses short-lived signed links for processing, blocks unsigned reads, and automatically expires website uploads.
Website uploads are configured to expire after one day, and normalized website result files after seven days. Operational and account data can have different retention needs as described in the privacy policy.
No. Customer files are processed to provide OCR output and are not used to train models.
The website API validates that upload keys and OCR job identifiers belong to the signed-in session. Cross-user upload and job references are rejected.
Yes. The website offers limited trial usage, and the Google Workspace add-on can be installed from Marketplace. Current plan details are listed on the pricing page.
Simple files often complete quickly, while large, multi-page, or complex documents take longer. Processing time also depends on current demand and the amount of layout analysis required.
Visit support and include the file type, approximate size, output mode, time of the issue, and exact error message. Do not email sensitive documents unless support specifically requests a secure transfer.
Follow detailed PDF, invoice, table, and scan-quality workflows.
Check scan quality and prepare spreadsheet column templates.
Get help with installation, account access, or OCR errors.