What PDF accessibility remediation actually costs
Published vendor rates run from about $5 to $25 per page, and $40 to $60 for scanned or complex documents. Here is what sits behind those numbers, and how to work out your own budget before you ask anyone for a quote.
Want your own number first? Run your PDFs through the free checker — it estimates the manual remediation cost for each file based on what it finds. Nothing is uploaded.
The going rate
Remediation is almost always priced per page rather than per document, because effort scales with page count and complexity rather than file count. Rates published by vendors in this market cluster like this:
| Document type | Rate per page |
|---|---|
| Simple text documents, already digital | $4 – $12 |
| Standard business documents | $5 – $25 |
| Scanned documents needing OCR | $25 – $60 |
| Complex tables, forms, charts | $40 – $60+ |
For a sense of scale: a 40-page board report at $10 per page is $400. An archive of 2,000 legacy documents averaging 12 pages each, at the same rate, is $240,000. This is why organisations with large back catalogues tend to discover the problem as a budget crisis rather than a technical one.
Why it costs what it does
Tagging a PDF by hand takes roughly 20 to 90 minutes per page depending on complexity. The work is not typing — it is deciding. A person has to open the document in Acrobat or a specialist tool and:
- mark every element as a heading, paragraph, list, figure, or table
- put those elements in a reading order that makes sense when heard rather than seen
- write alternative text for every image that carries meaning
- identify header cells in every table so data cells can be associated with them
- set the document title, language, and viewer preferences
- test the result with an actual screen reader
What drives your price up
Scanned pages
A scanned document is an image of words. Before anything can be tagged, it has to go through OCR, and OCR output has to be corrected — proper nouns, numbers, and anything in an unusual font are routinely misread. This roughly doubles or triples the per-page cost, and it is the single biggest factor.
Tables
Tables are the hardest structure to tag correctly, because the software cannot reliably tell which cells are headers. Nested tables, merged cells, and tables that span pages all require human judgement. Any vendor quoting a flat rate without looking at your documents is guessing.
Forms
Every field needs a label, a tooltip, and a sensible tab order. Interactive forms are usually priced separately and higher.
Charts and infographics
A chart needs alternative text that conveys the information, not the appearance. “Bar chart” is useless; “Enrolment fell from 4,200 in 2021 to 3,850 in 2025” is the actual content. Writing these well takes real time and real understanding of the data.
How to estimate your own budget
- Count your documents and total pages. Most people underestimate by a wide margin.
- Separate scanned documents from digital ones — price them differently.
- Flag anything with heavy tables, forms, or charts as complex.
- Apply the rates above, then add contingency. Legacy archives always contain surprises.
The free checker will do the first three steps for you across a whole folder at once, and tell you which documents are scanned.
Ways to spend less
- Fix the source, not the PDF. If you still have the Word or InDesign original, adding structure there and re-exporting is far cheaper than remediating the exported PDF. It also prevents the next version from having the same problem.
- Triage by traffic. Not every document deserves the same investment. Start with what people actually open.
- Consider whether it should be a PDF at all. An accessible web page is usually cheaper to produce and easier to maintain than an accessible PDF.
- Stop the bleeding first. Fixing the archive is pointless if new inaccessible documents are published every week. Train the people producing them.
A note on automated tools
Automated remediation — including ours, when we offer it — can reliably handle the mechanical parts: setting language and title, generating a structure tree, running OCR, drafting alternative text for review. It cannot reliably judge reading order or decide which cells are headers.
Anyone selling you fully automatic compliance is overselling. The honest version is that automation removes most of the cost and a human confirms the rest, which is a much better deal than paying a person to do all of it by hand.
Check your documents first
Before requesting quotes, it is worth knowing what you actually have. Run your PDFs through the free checker to see which are untagged, which are scanned, and roughly what fixing them would cost. Your files stay on your computer.