Audio & video
Audio Has No Text Version for People Who Can't Hear
A podcast or audio clip with no transcript or captions leaves people who are Deaf or hard of hearing with nothing at all.
audio-captionWCAG 2.0Introduced in WCAG 2.0.Level AA — the minimum. Missing it blocks people outright.
- Status
- CriticalCritical — Someone can complete the task only with great difficulty or outside help.Retiredaxe-core no longer runs this rule. Its card says why, and what still applies.
- Owner
- DevContentTestProduct
- Audit
- Automated+ManualA scanner finds it; a person confirms the fix.
Retired rule. Retired as an automated check: a scanner can't tell whether audio has a complete, accurate text alternative. WCAG 1.2.1 still applies — check it by hand.
Why it matters
Audio-only content, like a podcast, interview or voice message, holds all its meaning in sound. People who can't hear it get none of that meaning unless there is a text version.
The scanner looks for a captions <track> on the <audio> element. Most browsers don't show captions for audio, so the reliable fix is a full transcript on the page. axe-core has deprecated (retired) this rule for that reason, but the need is still real.
Who it affects
| Who | What they experience | Impact |
|---|---|---|
| Deaf and hard of hearing | They see a play button but get no words, so the whole clip is lost to them. | High impact |
| Cognitive and neurodivergent users | Without a transcript they can't read at their own pace, reread or search for a key point. | Partial impact |
Business & product impact
| Area | Why | Impact |
|---|---|---|
| Legal | Prerecorded audio without a text alternative fails 1.2.1 at Level A, the minimum. It also fails EN 301 549 and invites complaints under Australia's Disability Discrimination Act. | High impact |
| Conversion | If a product demo or offer is only in audio, some customers never hear it. | Moderate impact |
| SEO | A transcript adds real text that search engines can index. That is a genuine but small benefit. | Moderate impact |
| Brand trust | Posting podcasts with no transcript tells Deaf audiences they weren't considered. | Moderate impact |
| Cost to fix | Transcripts cost time or a small service fee per episode. Build it into the publishing workflow, not as a fix later. | Moderate impact |
Fixing it in code
A podcast episode on a page
Don't
<h2>Episode 12: Starting out</h2>
<audio controls src="ep12.mp3"></audio>A screen reader says: “Audio player — no words available to read.”
Do
<h2>Episode 12: Starting out</h2>
<audio controls src="ep12.mp3"></audio>
<h3>Transcript</h3>
<p><b>Mia:</b> Welcome to episode 12…</p>A screen reader says: “Audio player, followed by a heading "Transcript" and the full text.”
A transcript on the page works in every browser, unlike captions on audio.
Fixing it in design and content
Writing the transcript
Don't
A short summary: "In this episode we chat about getting started."
Do
Every spoken word, with who is speaking, plus key sounds like [laughter] or [music fades].
A summary drops detail; a full transcript gives people the same content.
Designer's handoff note
Place the transcript right next to the player, or link to it clearly with words like "Read the transcript".
Design a transcript layout that shows speaker names and is easy to scan.
Quick test
Tool
Your eyes, with the sound muted
Duration
1 minChecklist
- Pass
- Every audio clip has a full transcript you can find easily.
- Fail
- Any clip has no transcript, or only a short summary.
Check these at the same time
- A link has no text a screen reader can read
link-name - Sound plays by itself and can't be stopped quickly
no-autoplay-audio - A video has no captions
video-caption
What a green scan still misses
A scanner can only check for a <track> tag, not whether a transcript exists. It can't judge whether the text is complete or accurate — a person must read it.
Reference
- Status
- RetiredRetired as an automated check: a scanner can't tell whether audio has a complete, accurate text alternative. WCAG 1.2.1 still applies — check it by hand.
- Rule ID
audio-caption- Raised by
- axe-core: “<audio> elements must have a captions track”
- WCAG
- 1.2.1 Audio-only and Video-only (Prerecorded) (opens in a new tab) · Level A · since WCAG 2.0
- Other standards
- EN 301 549 9.1.2.1
- Last reviewed
- 16 October 2026