Captioning services compared on what the caption file carries

What WCAG asks a caption file to carry

The words are only part of it. W3C's Understanding note for 1.2.2 says captions “not only include dialogue, but identify who is speaking and include non-speech information conveyed through sound, including meaningful sound effects.” A file with every word right and no speaker labels can still leave a deaf viewer guessing who said the price, or missing the alarm that explains why everyone left the room.

That is the gap a service closes. W3C's guidance on making captions is blunt: automatically generated captions “do not meet user needs or accessibility requirements, unless they are confirmed to be fully accurate”, though they can be a starting point. Every service on this page either has people write the captions or has people correct a machine draft. What WCAG asks of recorded and live video, including transcripts, is set out in our guide to video caption requirements.

The order level that drops sounds and speakers

Every service in the table says its captions can carry speaker IDs and sound tags. Whether yours will depends on the level you order. Aberdeen's web captioning page is the most candid about it: its Basic level has no sound effects and no speaker IDs, and its Enhanced level has both. Verbit lists offscreen speaker identification as an add-on in its Captivate Post Plus tier. Both are fair products for other jobs; for 1.2.2, order the level that includes both, and say so in writing.

Before you sign, ask for a sample file from a video like yours, at the level you plan to order, and read it. The caption file check on our captioning software page flags missing sound tags, unnamed speakers and timing faults in a pasted SRT or WebVTT file. It cannot tell you whether the words are right; only watching with the sound on does that.

Order in time to watch the file first

Three of the five state a turnaround in plain numbers: Rev 12 hours or less, GoTranscript from 5 days down to a 6 to 12 hour rush, Aberdeen 24 or 48 hours on its Enhanced web level and 3 to 5 business days as typical. The number that matters is the day you must send the video so the captions come back with time to watch them before launch.

Which caption file your video host takes

A caption service hands back a file, and the file has to suit the player. W3C notes that WebVTT is the most common caption format on the web, with SRT and TTML the other usual ones. Name the format in the order rather than fix it after delivery.

If the video sits in an embedded player you do not control, check that the player shows captions and can be worked from a keyboard; the ADA website checklist has a video line for that, and our screen reader testing guide shows how to try the player's buttons yourself.

Live captions: CART, steno and what 1.2.4 covers

1.2.4 is about live media you broadcast: a streamed launch, a webinar, a council meeting on your site. W3C's Understanding note for 1.2.4 says it was meant for broadcast and is not meant to require captions on two-way calls between people through web apps. Its own example of live captions is CART, Communication Access Realtime Translation, where a trained writer turns speech into text as it happens.

Three services here state a human live option. Aberdeen's live captioners write on a stenographic keyboard, and it guarantees 200 words a minute at 98% accuracy or higher. 3Play Media says its professional live events are captioned by experienced people rather than auto-generated. Verbit sells CART for a room or a remote audience and states 24-hour booking. All three also sell automatic live captions, so name the human option in the booking.

Public bodies have a dated reason to care. The DOJ Title II rule's technical standard is WCAG 2.1 Level AA, and 1.2.4 is a Level AA criterion. Governments with a population of 50,000 or more face 26 April 2027; smaller ones and all special district governments, 26 April 2028, after DOJ's April 2026 interim rule pushed each date back one year. Check ada.gov for the current dates; the detail is in our Title II web rule guide.

Hire a service or caption it yourself

A service is the right buy when the videos are long, have several speakers, use trade terms, or keep coming. For two short clips with one clear speaker, editing a machine draft yourself is often enough, as long as someone adds the speaker labels and sounds and checks every line. The tools for that, from caption editors to export formats, are in closed captioning software compared.

If you hire, send three things with the order: the speakers' names as they should appear, a list of product names and trade words with their spelling, and the file format your player takes. 3Play Media calls this a cheatsheet and Verbit calls it prep material; any captioner works faster and better with it.

Captions cover what is heard. What is only seen, such as a price on screen or a step shown and not said, is the job of audio description: 1.2.3 at Level A asks for audio description or a text alternative, and 1.2.5 at Level AA asks for audio description. Three of these five sell it; the specialists are in audio description services compared, and what audio description is explains when a video needs it.