Caption compliance

Prove the state of your captions, asset by asset

ADA Title II web rule — large public entities names WCAG 2.1 Level AA, and takes effect 26 April 2027 — 7 months away. When a reviewer asks what you can show for it, the answer has to be evidence about specific files on a specific date — not a policy document.

Talk to us about Compliance See pricing

The clock

European Accessibility Act
In force since 28 June 2025
Businesses selling covered products and services into the EU, including audiovisual media services.
Technical standard: EN 301 549 V3.2.1 · Directive (EU) 2019/882
ADA Title II web rule — large public entities
Takes effect 26 April 2027 — 7 months away
US state and local government entities serving a population of 50,000 or more, including public universities and school districts.
Technical standard: WCAG 2.1 Level AA · 28 CFR Part 35, DOJ final rule
ADA Title II web rule — small entities
Takes effect 26 April 2028 — 19 months away
US state and local government entities serving fewer than 50,000 people, and special district governments.
Technical standard: WCAG 2.1 Level AA · 28 CFR Part 35, DOJ final rule

What we cannot tell you

CaptionPass reads caption files. It never hears the audio. Everything below is a consequence of that, and it is on every receipt and every report we produce — not in a footnote.

We cannot confirm the captions match what was said.: That comparison needs the audio, and the engine only ever sees the caption file. Accuracy under FCC 79.1(j)(2)(i) is reported as not evaluated on every receipt we issue — permanently, by design. Any vendor telling you otherwise from a sidecar file is telling you something they cannot know.
We cannot answer anything about live captioning.: WCAG 1.2.4 is a property of the broadcast, not of a delivered file. It is listed on the report so its absence is visible, and marked not evaluated rather than Not Applicable — because it is entirely applicable to anyone who streams.
We can only partly answer placement.: Line count and dropped positioning are visible in the file. Whether a caption covers a lower third or a scoreboard needs the video, so the report says which half it checked.
A clean scan is not a conformance proof.: Most caption criteria can be disproved from a file but never proved. Those criteria report as Partially Supports even when nothing is wrong, rather than being rounded up to Supports. It is the row a careful reviewer looks for first.

What we can

A signed receipt for every delivery

Each delivery produces a conformance receipt: the standard it was graded against, the verdict on each clause, the findings behind a failure with the cue they came from, and the engine version that decided it. Signed, so a reviewer can tell whether anything changed after it was issued.

Look at a real one →

Anyone you send it to can check it

Every receipt has a public URL. No account, no login — the reader sees what was checked, what was not, and whether the document is byte-for-byte what was issued. The signature travels with the file, so a receipt still verifies if we are not the ones holding it.

An Accessibility Conformance Report across the library

The same evidence rolled up: every caption criterion of a named standard, answered across every asset, with the failing assets named under each row. Printable HTML for the RFP response, and CSV in the column order a VPAT uses.

Read a real one →

We publish our own report, including what it fails

CaptionPass has its own Accessibility Conformance Report, generated from a repeatable audit of the application and published after the audit ran rather than before. One criterion is currently Does Not Support, with the count and the remediation plan on the page. A vendor asking you to trust their evidence should be willing to be measured by it.

Read our conformance report →

Four standards, cited clause by clause

WCAG 2.1 Level AA, EN 301 549 V3.2.1, FCC 47 CFR 79.1(j)(2), and the DCMP Captioning Key. Each clause quotes the requirement in the standard's own words and links the document, so a reviewer can check the citation rather than take it on trust.

See the standards and what each one covers →

Who this is for

University media and disability services: A dated obligation across a library nobody has fully inventoried, and a procurement team that will ask for an ACR.
Government and agency communications: Public-facing video under Title II, and a records requirement that outlives whoever captioned it.
Vendors delivering into public institutions: An RFP response that has to answer caption criteria without a blank row or an unjustified Not Applicable.

If you are a freelance editor fixing one file before a deadline, the free converter is the page you want — it does that job in one upload, with no account.

Compliance

$99 / month

200 conformance receipts included, then $0.50 each. No seat requirement — the person who needs the evidence is usually not the person editing captions.

Start a conversation

Billing opens soon — the waitlist gets the founding rate.

Questions we get asked

Does CaptionPass make our media accessible?: No — it establishes what is true about your caption files and gives you evidence you can hand to somebody else. Captioning the media is still your work or your vendor's. What changes is that you can prove the state of it, per asset, on a date, with the engine version that decided.
Can we use a receipt in an accessibility conformance report?: That is what it is for. The library-level report is generated in the same format procurement expects, with per-asset evidence behind each row, and every criterion the engine cannot settle from a caption file says so with the reason rather than being left blank.
What happens to a receipt if we stop being a customer?: It keeps verifying. The signature travels with the JSON, so anyone holding the file can confirm it is unaltered without us being involved — the hosted verify page is a convenience, not the proof.
Do we need seats for everyone who needs the reports?: No. The Compliance plan has no seat requirement, because the person who needs conformance evidence for a library is usually not the person editing captions. Buying editing seats to reach a report would be paying for the wrong thing.
How current are the standards you cite?: Each profile names the document and its version — WCAG 2.1 (W3C Recommendation, 5 June 2018), EN 301 549 V3.2.1 (2021-03), 47 CFR 79.1(j)(2), and the DCMP Captioning Key. Receipts pin the engine version too, so a verdict can be reproduced or challenged rather than taken on trust.

Prove the state of your captions, asset by asset

The clock

European Accessibility Act

In force since 28 June 2025

Businesses selling covered products and services into the EU, including audiovisual media services.

Technical standard: EN 301 549 V3.2.1 · Directive (EU) 2019/882

ADA Title II web rule — large public entities

Takes effect 26 April 2027 — 7 months away

US state and local government entities serving a population of 50,000 or more, including public universities and school districts.

Technical standard: WCAG 2.1 Level AA · 28 CFR Part 35, DOJ final rule

ADA Title II web rule — small entities

Takes effect 26 April 2028 — 19 months away

US state and local government entities serving fewer than 50,000 people, and special district governments.

Technical standard: WCAG 2.1 Level AA · 28 CFR Part 35, DOJ final rule

What we cannot tell you

CaptionPass reads caption files. It never hears the audio. Everything below is a consequence of that, and it is on every receipt and every report we produce — not in a footnote.

We cannot confirm the captions match what was said.

That comparison needs the audio, and the engine only ever sees the caption file. Accuracy under FCC 79.1(j)(2)(i) is reported as not evaluated on every receipt we issue — permanently, by design. Any vendor telling you otherwise from a sidecar file is telling you something they cannot know.

We cannot answer anything about live captioning.

WCAG 1.2.4 is a property of the broadcast, not of a delivered file. It is listed on the report so its absence is visible, and marked not evaluated rather than Not Applicable — because it is entirely applicable to anyone who streams.

We can only partly answer placement.

Line count and dropped positioning are visible in the file. Whether a caption covers a lower third or a scoreboard needs the video, so the report says which half it checked.

A clean scan is not a conformance proof.

Most caption criteria can be disproved from a file but never proved. Those criteria report as Partially Supports even when nothing is wrong, rather than being rounded up to Supports. It is the row a careful reviewer looks for first.

What we can

A signed receipt for every delivery

Look at a real one →

Anyone you send it to can check it

Who this is for

University media and disability services

A dated obligation across a library nobody has fully inventoried, and a procurement team that will ask for an ACR.

Government and agency communications

Public-facing video under Title II, and a records requirement that outlives whoever captioned it.

Vendors delivering into public institutions

An RFP response that has to answer caption criteria without a blank row or an unjustified Not Applicable.

If you are a freelance editor fixing one file before a deadline, the free converter is the page you want — it does that job in one upload, with no account.

Questions we get asked

Does CaptionPass make our media accessible?

No — it establishes what is true about your caption files and gives you evidence you can hand to somebody else. Captioning the media is still your work or your vendor's. What changes is that you can prove the state of it, per asset, on a date, with the engine version that decided.

Can we use a receipt in an accessibility conformance report?

That is what it is for. The library-level report is generated in the same format procurement expects, with per-asset evidence behind each row, and every criterion the engine cannot settle from a caption file says so with the reason rather than being left blank.

What happens to a receipt if we stop being a customer?

It keeps verifying. The signature travels with the JSON, so anyone holding the file can confirm it is unaltered without us being involved — the hosted verify page is a convenience, not the proof.

Do we need seats for everyone who needs the reports?

No. The Compliance plan has no seat requirement, because the person who needs conformance evidence for a library is usually not the person editing captions. Buying editing seats to reach a report would be paying for the wrong thing.

How current are the standards you cite?

Each profile names the document and its version — WCAG 2.1 (W3C Recommendation, 5 June 2018), EN 301 549 V3.2.1 (2021-03), 47 CFR 79.1(j)(2), and the DCMP Captioning Key. Receipts pin the engine version too, so a verdict can be reproduced or challenged rather than taken on trust.