Sign in Request an invite

PDF accessibility remediation — verifiable WCAG and PDF/UA-1 outcomes

2,562,231
32

verapdf · anchor corpus · 268 pdfs · 262 passing

Built on the pipeline that processed 35,000+ PDFs for federal court clients.

Most PDFs on government and enterprise websites are visually fine but structurally empty. To assistive technology they are a flat image.

We rebuild the structure so screen readers, voice control, and refreshable braille can actually use the document — and we prove the result with veraPDF before it ships.

Proof: anchor corpus

full proof page

A real run on a real public-sector corpus, validated end-to-end with veraPDF.

documents passing wcag 2.2
262 of 268
wcag rule violations remaining
2,562,231 32
pdf/ua-1 rule violations remaining
2,572,599 32
documents passing pdf/ua-1
262 of 268

40 corpora across ten industries — international standards bodies, federal agencies, law firms, insurance carriers, government affairs firms, think tanks, trade associations, public utilities, court systems, state education boards.

Anchor case study: customer attribution forthcoming. Aggregate corpus stats appear anonymized on the proof page. Other corpora are remediated under existing engagements; per-customer case studies require sign-off and are not published on the public site.

Every run ends with a veraPDF report so the WCAG 2.2 and PDF/UA-1 violation deltas are checkable, not asserted.

Anchor corpus
268 PDFs · 97.8%
w3.org
4 PDFs · 4 of 4 fixed
access-board.gov
159 PDFs · 139 of 159 fixed
ada.gov
618 PDFs · 506 fixed
section508.gov
4 PDFs · 4 of 4 fixed
Policy think tank
35,400 PDFs · 97.4%

The math, also

The industry charges per page. We charge per PDF.

$12,648
corpus total at ~$4 / page (industry)
$1,340
corpus total at $5 / pdf (Project tier)

Same anchor corpus. 3,162 pages across 268 PDFs (average 11.8 pages per PDF).

industry per-page rate: ~$4/page low end · published quotes for pdf accessibility remediation, 2025

Or compare both numbers to the price of doing nothing →

How it works

A deterministic pipeline. Same source PDF in, same remediated PDF out, with a veraPDF report attached.

  1. 01

    Extract

    Pull text, structure cues, and metadata out of the source PDF.

    bytes → structured input
  2. 02

    Structure

    Rebuild the logical document tree so headings, lists, tables, and reading order are explicit.

    tags: 0 → 247
  3. 03

    Glyph mapping

    Repair character mappings so assistive technology can read glyphs back as text.

    tounicode: missing → present
  4. 04

    Identification

    Stamp title, language, and the PDF/UA-1 conformance identifier — the only conformance marker PDFs natively carry.

    xmp: missing → present
  5. 05

    Annotations

    Assign required accessibility roles to every interactive element — links, form fields, alt-text slots — so each is reachable.

    roles: 0 → 18
  6. 06

    Validation

    Validate against PDF/UA-1 and WCAG 2.2 with veraPDF and emit the violation delta.

    violations: 251 → 2

What your first run will look like

verapdf report preview

A faithfully-mocked veraPDF delta in the product's own treatment. Every real run lands a report shaped like this against your actual corpus.

wcag 2.2

247 0

pdf/ua-1

251 2

sample delta · single document · anchor corpus