Video Accessibility Guide 2026: Captions, Transcripts & Audio Descriptions

Video without captions excludes an estimated 466 million people with disabling hearing loss worldwide. WCAG 2.1 devotes an entire guideline (1.2 — Time-based Media) to video and audio accessibility. This guide covers every requirement, what qualifies as compliance, and how to implement captions, audio descriptions, and transcripts correctly.

Captions: Requirements and Quality Standards

Captions are synchronized text that appears on screen as video plays, conveying all spoken dialogue, speaker identification, and meaningful non-speech audio (e.g., [applause], [phone ringing], [upbeat music]).

What WCAG Requires for Captions

WCAG 1.2.2 requires captions for all prerecorded synchronized media (video + audio). The captions must be:

• Synchronized — timed to match the spoken audio within ~2 seconds of when it's spoken.
• Equivalent — conveying all speech, speaker identification (when multiple speakers), and relevant non-speech audio.
• Accurate — WCAG doesn't define an explicit accuracy percentage, but legal enforcement and best practice guidelines (including the FCC's broadcast standards) use 99%+ accuracy as the threshold.
• Accessible — the caption mechanism itself must be keyboard-operable and not require a mouse to enable.

Open vs. Closed Captions

Open captions are burned into the video — always visible, can't be turned off. They satisfy WCAG unconditionally but may not match all user styling preferences.

Closed captions are delivered as a separate text track (WebVTT, SRT, or TTML file) that users can enable, style, and position. Both satisfy WCAG 1.2.2 — closed captions are preferred because users can customize display.

Auto-Captions: A Starting Point, Not a Solution

YouTube, Vimeo, Microsoft Teams, and Zoom all offer automatic speech recognition (ASR) captions. ASR accuracy averages 80–90% in ideal conditions and drops sharply for:

• Technical jargon, product names, and acronyms
• Non-native speaker accents
• Multiple overlapping speakers
• Background noise or music
• Low-quality audio recordings

An 85% accuracy rate means one in six words is wrong. For a legal brief, medical explanation, or technical tutorial, that's not accessible. Auto captions must be reviewed and corrected before being published.

Audio Descriptions

Audio descriptions provide narration for visual content that isn't described in the existing audio track. They're inserted during natural pauses in dialogue to describe what's happening visually — actions, scene changes, on-screen text, expressions, and other visual information necessary to understand the content.

Audio descriptions primarily serve blind and low-vision users who can't see the video. Without them, a video tutorial that shows a cursor moving and clicking — without narrating each step — is largely unintelligible through audio alone.

When Are Audio Descriptions Required?

WCAG 1.2.5 (Level AA) requires audio descriptions for all prerecorded video with visual information essential to understanding the content. The exception: if all important visual information is already narrated in the audio track ("I'm now clicking the Settings menu in the top-right corner"), audio descriptions are redundant and not required.

Common video types that require audio descriptions:

• Product demos and software tutorials showing screen content
• Training videos with visual demonstrations
• Marketing videos with on-screen text, charts, or infographics
• Conference presentations showing slides
• Video testimonials where speaker expressions convey meaning

Creating Audio Descriptions

There are two main approaches:

• Standard audio description — Descriptive narration is recorded and inserted during natural pauses in the original audio. Requires that pauses be long enough to describe the visual content.
• Extended audio description (WCAG 1.2.7, Level AAA) — When pauses aren't long enough, the video pauses entirely to allow longer descriptions. Required for very dense visual content with minimal silence.

For new video production: the most efficient approach is to write scripts that fully describe actions verbally — "I'll click the blue Submit button at the bottom of the form" — eliminating the need for audio descriptions entirely.

Transcripts

A transcript is a text document containing all spoken content from a video or audio file. Unlike captions, transcripts aren't synchronized — they're a separate document that can be read independently.

Transcripts serve several audiences:

• Deaf-blind users who use refreshable Braille displays
• Users who prefer to skim or read rather than watch a video (especially for instructional content)
• Search engines — transcripts are indexable text that improves SEO
• Users in loud environments or who can't use audio at the moment

What Must a Transcript Include?

For a transcript to satisfy WCAG 1.2.3 (Level A) and WCAG 1.2.8 (Level AAA), it must be a full alternative — meaning it includes all:

• Spoken dialogue (verbatim or clean read, either is acceptable)
• Speaker identification when multiple speakers
• Non-speech audio that conveys meaning: [laughter], [applause], [alarm sounds], [dramatic music]
• Visual information (for video transcripts): descriptions of on-screen actions, charts, slide content, and visuals not described verbally

Accessible Media Players

Even with captions and audio descriptions, your video is inaccessible if the media player itself can't be operated by keyboard or screen reader. WCAG requires that all functionality of the media player — play, pause, volume, seek, caption toggle, fullscreen — be keyboard-accessible.

Media Player Accessibility Requirements

Accessible Video Players

Not all embedded video players are accessible. YouTube's embedded player is generally keyboard-accessible but has variable caption toggle accessibility depending on embed parameters. Third-party players like Able Player, Plyr, and Video.js with proper configuration offer better accessibility support.

For YouTube embeds, use ?cc_load_policy=1 in the embed URL to force captions on by default, and ensure captions are available and accurate before embedding.

Video Accessibility Checklist

Frequently Asked Questions