Free tools
Six free browser tools that answer the questions a remediation decision starts with: how many PDFs does the website publish, what does a file look like as a web page, is it tagged, what do the tags say, can tags be added without redrawing the page, and is there any real text under the scan. No account, and uploads are processed for the request rather than retained.
Count
Enter a domain and get every public PDF the site links to, the pages linking to each one, and an automated check on the most-linked files. Nothing to install.
Walks the whole site, opens the 10 most-linked files, no account
Open website pdf scannerConvert
Turn a PDF into a real web page and see the result immediately: headings, lists, tables with header cells, figures, and a reading order a screen reader can follow.
First 10 pages, 2 documents per day, no account
Open pdf to html converterDiagnose
Audit tags, metadata, figures, tables, lists, forms, and PDF/UA machine rules. Automated evidence stays separate from the checks a person still has to make.
Returns machine findings plus an explicit human-review checklist
Open pdf accessibility checkerInspect
Read the logical tag tree of any PDF in the browser: headings, tables, figures, role maps, and per-tag assessments. Read-only, and no Acrobat Pro required.
Read-only inspection, nothing is written back to the file
Open pdf tag viewerRepair
Auto-tag an eligible born-digital PDF, inspect the exact candidate tree, correct paragraph and heading roles, and download the reverified result without redrawing a page.
5 upload attempts per UTC day, up to 20 pages each, no account
Open pdf tag editorRecover
Detect scanned and OCR-derived pages and route them out of strict automatic tagging before recognition errors can become confident-looking structure.
Per-page scan/OCR eligibility check, fail closed
Open scanned pdf accessibilityWhich tool first
Most PDF accessibility work goes wrong at the first step, because the file was never inspected. A document can look finished and still carry no tag tree at all, or carry a tag tree that describes a different reading order than the one on the page. Start with the PDF accessibility checker to see what the machine can prove, then open the tag viewer to read what the structure actually claims.
If the file has no usable text under the image, tagging cannot help until the text is recovered. Run scanned PDF recovery first in that case. If the text is present but the structure is missing, the auto-tagger adds a logical structure without altering the page as drawn.
These tools report machine-detectable conditions. A clean automated result is not a conformance claim, and it does not replace testing with a screen reader. The PDF accessibility checklist sets out the manual checks that remain, and PDF/UA vs WCAG explains which standard each finding belongs to.
When a whole backlog needs the same treatment rather than one file at a time, compare the remediation paths before committing: hosted accessible HTML, a rebuilt tagged PDF, and manual exact-layout remediation solve different problems and cost different amounts of work.
The free tools answer questions about a single document. When the same questions apply to hundreds of files across several departments, the work needs an inventory, an owner per document, a review step, and a record of what was published when.
See the full workflow