Purpose
- Give public buyers a comparison they can rely on, from a tester that is not the vendor.
- Give tool makers and vendors an independent result they can cite, good or bad.
How it works
- Documents: real public-body documents across text, tables, forms and scans.
- Held-out set: the test documents are not shared with entrants before testing.
- Checks: PDF/UA validation, no loss or change of text compared with the original, reading order, structure quality, and human spot checks.
- Vendor first: each entrant sees its own results before anything is published.
- Anonymous unless you consent: commercial entrants appear as "Subject A, B, ..." unless they agree in writing to be named. Results are measurements of the named versions only, not endorsements and not accessibility or legal determinations.
- Tools can fail: a tool that does not meet the criteria is reported as not meeting them.
The criteria are the public PDF acceptance criteria.
Disclosure
Trilatic also provides remediation services. Our own output is not ranked in this benchmark.
Enter your tool or service
Inclusion in the pilot is free for vendors and tool makers.
Opens your email app, addressed to contract@trilatic.com.
Timeline
Pilot results: Q4 2026. No results have been published yet.
Results describe the tested documents and are not a compliance determination. Not legal advice.