Help & troubleshooting

Check that the file is a real, non-empty PDF, not another format renamed to .pdf. Password-protected PDFs are unsupported. Open your own file with its password in the original application and export an unencrypted copy. If the file is damaged, download or export a fresh copy.

PDF tools accept up to 25 MB per file and 50 pages per task. Merge accepts up to 20 PDFs and 50 MB combined, still within 50 pages total. Image tasks accept JPG, PNG or WebP up to 25 MB and 40 megapixels per image; Photos to PDF accepts up to 20 images. Split a large source in its original application before choosing it here.

PDF to text reads existing text; it does not recognize a photograph of words. Try selecting a sentence in your PDF viewer. If the page is a scan, open PDF editor and choose Read a scan. A document can contain both scanned and digital pages.

Use a clear, upright scan and choose the appropriate OCR language. Recognition is not a transcription guarantee: compare names, dates, punctuation and similar letters with the original. Process one page at a time and correct the recognized text before using it. Handwriting and complex tables may be unreliable.

Optimization may save little on an already efficient PDF. If the result is larger, the tool keeps the original for download. For a scan, try image-based compression after reading its notice. This removes selectable text, clickable links and interactive fields. Compare small print before accepting the smaller copy.

A very small KB target can reduce detail. Increase the size limit or reduce dimensions only as far as the destination allows. Maximum width and height preserve proportions: they do not crop an image to an exact square. Check the downloaded copy with File checker. Animated images are not preserved as animations by these image workflows.

Merge follows the file list shown in the tool. Extract pages follows the numbers you enter, such as 3, 1-2 or ٣، ١-٢. Positions start at the first PDF page, even if the printed numbering starts later. Repeated page numbers are included only once. Use the practice guide to compare a known result.

Supported form fields need a saved appearance. Open the source in its original PDF viewer, save a reviewed copy and retry. XFA forms are unsupported. Keep an editable original because flattening fixes the fields. A drawn signature is not a certificate-based digital signature, and editing a signed PDF may invalidate its signature.

Look in your browser download list and the device folder you selected. On a phone, the Files app may show a Downloads folder. Open the downloaded result rather than the original. A preview tab alone does not guarantee a saved copy. Use the tool’s Download button and confirm that a file appears on your device.

Keep the tab open during processing. For a large task, try fewer pages or lower image resolution and close unused tabs. If the embedded tool does not load, use its Open the full workspace link. Download any existing result before reloading; reloading can clear work. The first OCR use also needs to load local site assets over the internet.

The current tools process selected documents in your browser; they do not upload the contents to a conversion server. Results are temporary until you download them. Site delivery still makes ordinary network requests to the hosting provider. Sharing a page link does not include your files. Read the privacy notice for the distinction.

Not in this version. Office documents, audio, video, HEIC, SVG and archive uploads are not supported conversion inputs. TXT and ZIP are outputs of certain tools, not general input converters. Changing a filename extension does not convert its contents. See the supported-format table before starting.