All 22 tested workflows produced valid outputs that matched the requested operation for their named stress case. That does not mean every possible PDF is supported; it means each published result below has direct output evidence rather than a page-load check.
Results by workflow
The checksum identifies the exact downloaded artifact used for the recorded result. Word conversion rows were re-run after the v81 reconstruction upgrade; the remaining rows come from the full real-output release audit completed on the same fixture corpus.
| Workflow | Stress case | Independently verified result |
|---|---|---|
| Merge PDF | 3 PDFs, 18 pages, reordered on mobile | 18 pages; exact reordered marker sequence preserved |
| Split PDF | 100-page PDF split every page | 100 valid one-page PDFs; markers 1, 50, and 100 preserved |
| Compress PDF | 9.2 MB four-page image PDF, Maximum preset | 516 KB, four valid pages, visually correct; 94.4% smaller |
| Rotate PDF | Pages initially at 0°, 90°, 180°, and 270° | Output rotations 90°, 90°, 180°, and 270° as requested |
| Organize PDF | Move page 2 first; delete page 3 | 11 pages; marker order 2, 1, 4 through 12 |
| Delete Pages | Delete pages 3 and 7 | 10 pages; only the requested markers absent |
| Extract Pages | Extract pages 2, 5, and 9 | Three pages in the exact requested order |
| Sign PDF | Typed signature on page 5, top-left | Signature image appears only on page 5 at the requested position |
| Watermark PDF | Red 50% watermark on pages 2 and 4 | Text exists only on pages 2 and 4; render is visually correct |
| Page Numbers | Pages 2-12, starting at 10, “Page n of N” | Page 1 unchanged; pages 2-12 labeled 10 through 20 |
| Crop PDF | Mixed sizes; crop pages 2-4 by 10 mm | CropBox changed by 28.35 points only on pages 2-4 |
| Protect PDF | 12 pages, strong password | Blocked without password; every page readable with it |
| Unlock PDF | Wrong password, then correct password | Friendly rejection, followed by lossless unlocked 12-page output |
| PDF to Text | Pages 2, 5, and 9 with separators | 8,608 characters; only requested page markers present |
| OCR PDF | Image-only scan | Unique marker and numbered lines recognized in TXT and valid DOCX |
| PDF to Word | Three pages, real table, image, columns, headers, footers, landscape | Three pages; landscape geometry, numbered lists, headers, footers, and left-before-right order preserved |
| Word to PDF | DOCX with headings, styles, lists, merged table, image, headers, footers | Searchable Letter PDF; geometry, constrained image, and basic header/footer preserved |
| JPG to PDF | 1.4 MB photo plus small JPG | Two correct pages; optimized output 814 KB |
| PNG to PDF | Transparent PNG | One valid page; transparency displays correctly |
| HEIC to PDF | Real HEIC photo container | One visually correct PDF page; 810 KB |
| PDF to JPG | Pages 2 and 5 at 144 DPI | Two 1,224 × 1,584 JPEG files in a valid ZIP |
| PDF to PNG | Pages 1 and 3 at 200 DPI | Two 1,700 × 2,200 opaque-white PNG files in a valid ZIP |
How the outputs were checked
- 1. Synthetic fixtures
Files contain unique page and content markers, deliberate rotations, transparency, images, columns, encryption, and known geometry. No private user documents are used.
- 2. Real browser workflows
Files are selected through the interface, controls are changed, processing finishes, and the actual download is captured.
- 3. Independent parsers
Outputs are read with pypdf, pdfinfo, python-docx, ZIP validation, and text extraction rather than trusting the code that created them.
- 4. Visual rendering
Poppler and ImageMagick renders catch clipping, black transparency, missing images, incorrect placement, and page-orientation errors that metadata alone misses.
- 5. Checksums
Every row in the CSV includes a SHA-256 hash for the exact output evaluated.
Important limits
This is a product validation dataset, not a market-wide competitor ranking. A passing row applies to the stated fixture and tested browser workflow. Browser memory, damaged files, unusual encodings, handwriting, certificate-backed signatures, PDF/A requirements, and highly designed Word layouts can require specialized desktop software.
The Word conversion tests are deliberately demanding because PDF and DOCX encode layout differently. The PDF to Word tool offers editable reconstruction and an appearance-preserving visual mode; the Word to PDF tool labels its browser-pagination limits directly.
How to cite or reuse the data
The CSV is licensed under CC BY 4.0. Cite “GoPDFConverter Browser PDF Tools Accuracy Benchmark 2026,” link to this page, and name the workflow and stress case you use. Do not generalize a fixture result into a guarantee for every document.