How PDFMK Processes Your Files
A precise walkthrough of upload, temporary processing, download, and cleanup for every PDFMK tool.
Published · Updated by the PDFMK Editorial Team
Start with the correct mental model
Merge, Split, Organize, common image-to-PDF jobs, and supported PDF-to-image jobs run on your device. Watermarks send only text and appearance settings to prepare a small asset, then process the PDF locally. If local processing cannot finish, files are automatically uploaded over HTTPS for temporary server processing, with a visible status and cancel option. PDF to Word uses temporary server processing. Selecting files generates local previews without transmitting the documents.
The processing status identifies whether a job runs on your device or the server. Local transformations retain native PDF text and graphics. If a file includes interactive or document-level features, the existing server engine remains available to preserve tool compatibility. If local processing cannot finish, the processing dialog announces the automatic server fallback and shows upload, processing, and download progress. You can cancel at any stage.
What each tool does behind the interface
The seven tools use different PDF and image libraries and therefore do not all handle temporary data in exactly the same way. The relevant question is not simply whether a server is involved, but what the server must create to complete the job.
| Tool | Working method | Returned download |
|---|---|---|
| Merge | Supported jobs run locally. If local processing cannot finish, files automatically use temporary server processing with a visible status and cancel option. | merged.pdf |
| Split | Supported jobs run locally. If local processing cannot finish, files automatically use temporary server processing with a visible status and cancel option. | One PDF or a ZIP |
| Organize | Supported jobs run locally. If local processing cannot finish, files automatically use temporary server processing with a visible status and cancel option. | organized.pdf |
| Images to PDF | Supported jobs run locally. If local processing cannot finish, files automatically use temporary server processing with a visible status and cancel option. | images.pdf |
| Watermark | Only watermark text and appearance settings go to the server; a small asset is applied to the PDF on your device. Compatibility failures use temporary full-document server processing. | watermarked.pdf |
| PDF to Word | The first-page preview and text-layer check stay local. Selected pages are converted in a temporary working directory after export starts. | document.docx |
| PDF to images | Supported PDFs and ZIP inputs render locally using a downloaded engine. Engine loading or compatibility failures use temporary server processing. | A ZIP of PNG files |
When cleanup happens
Each server operation uses a request-specific temporary working directory. Cleanup runs directly on validation or processing errors and after a successful file response finishes. Split previews use browser memory and object URLs that are released when you replace the file or leave the page. A closed tab or failed download does not give PDFMK a future retrieval feature; you normally need to run the operation again.
Operational network and security logs are a separate category from document working files. Hosting infrastructure can record request time, IP address, URL, browser information, response status, and abuse signals. Those records help keep the service available, but they should not contain the body of an uploaded PDF. A filename can appear in processing context or a support report, so avoid putting secrets in filenames.
A practical sensitivity check
Before uploading, ask two questions. First, are you authorized to send this document to an external online processor? Second, would a temporary server copy violate a law, contract, client promise, or internal policy? If either answer is uncertain, stop and use a managed offline tool or ask the document owner.
Documents that often need extra care include medical records, identity documents, bank statements, legal material under privilege, student records, unreleased financial information, and files covered by a data-processing agreement. Redaction is only useful when it removes the underlying PDF objects, not when it places an opaque rectangle over visible text. Verify redaction with a dedicated tool before using any online processor.
- Work from a duplicate and keep the original unchanged.
- Remove unnecessary pages before uploading when policy allows it.
- Open and inspect the downloaded file before deleting your local working copy.
- For highly sensitive material, choose an approved offline workflow instead of relying on cleanup promises.
What advertising can and cannot see
PDFMK uses Google AdSense on site pages. Advertising technology can receive ordinary page and browser data and may use cookies or identifiers according to region and consent choices. PDFMK does not intentionally send the content of an uploaded document to Google for ad targeting, and ads are not inserted into the PDF output.
Local operations keep document contents on the device. Server operations receive documents for the requested transformation, use temporary storage, and remove working artifacts through cleanup. Ordinary website requests and analytics remain separate from document processing. Read the Privacy Policy and Cookie Policy for details.