Real output checks
Core workflows are tested by creating or opening documents, calling production services and inspecting the resulting PDF or DOCX—not only by testing button labels or mocked return values.
Who, how and why
PDF software should earn trust with inspectable evidence
PDFCore is a Windows document-tool project focused on useful local workflows: opening and organizing PDFs, OCR, PDF-to-Word conversion, merging, splitting, compression and protection without an account or document upload.
Why PDFCore exists
PDF documents often contain private contracts, financial records, identification, research and internal business material. Many common tasks do not need a server. PDFCore is designed so the desktop application can perform its principal workflows on the Windows PC.
“Local” is not used as a vague privacy slogan. The current application has no account system, advertising, analytics, telemetry, web server or automatic document uploader. The website can be opened deliberately from the app, but opening, OCR, conversion and saving do not require sending the document to PDFCore.
Current release
PDFCore 8.0.1 for Windows x64
Supports Windows 10 and Windows 11, ten interface languages and a self-contained installed runtime. The public installer is free for personal and commercial document work.
The current installer is not Authenticode-signed. That limitation is disclosed on the download page rather than hidden behind a generic trust claim.
How claims are prepared
Product pages are tied to tests, artifacts and explicit limits
Artifact identity
The download page publishes version, file size and SHA-256. Sample benchmark files have their own manifest so readers can verify that the inspected files did not change.
No unsupported ranking
PDFCore is working toward stronger PDF-to-Word quality, but it does not claim a Top‑5 position without a diverse same-input blind comparison and independent review.
Who writes and reviews the technical content
Pages labeled as PDFCore Engineering are prepared from the product implementation, release tests, fixed sample artifacts and independent release-audit evidence. The organization is the named author because the current public material represents the maintained product record, not an individual opinion or a fabricated expert persona.
Where a guide explains a general PDF or Word limitation, it distinguishes that limitation from a PDFCore-specific feature. Where evidence is narrow—for example, a synthetic PDF-to-Word corpus—the page states the boundary and the next verification gate.
What we are improving next
- A 30+ document, rights-cleared PDF-to-Word benchmark with blind review
- Merged-cell and borderless-table reconstruction
- Mixed-script, rotated-text and real-scan coverage
- Code signing and broader clean-machine installation evidence
- Working, verified support operations and clearer public change records
Contact, privacy and corrections
Use the support page for product questions, reproducible defects and corrections to published evidence. Do not send confidential documents unless you have created a safe synthetic example. The privacy page describes desktop and website data handling, and the benchmark methodology explains how comparative evidence will be produced.
Inspect before you trust
Download the public test PDF and the three Word outputs, or verify the installer hash before running it.