Maxademics · Blog · Accessibility

How to Caption a Video for Canvas — Free + Paid Tools (2026)

Auto-captions get you to ~85% accuracy. WCAG and ADA Title II require ~99%. That last 14% takes about 3 minutes of editing per minute of video — but you can either do it yourself for free, or pay $1.50/minute to outsource it. Here's the full teacher's workflow for both paths in 2026.

Why captions are legally required (not optional)

If your institution receives federal funding, ADA Title II and Section 504 require video content to be accessible to students with hearing disabilities. WCAG 2.1 AA — the standard most universities are audited against — specifies captions must be accurate enough that a deaf student gets the same information as a hearing student. In practice, that's ~99% word accuracy.

The Office for Civil Rights has settled cases against multiple universities for posting uncaptioned video. The penalty is rarely a fine — it's institutional embarrassment, a mandatory remediation plan, and a permanent compliance scar. None of that lands on you personally; it lands on your department.

What "good captions" actually look like

WCAG defines captions as a synchronized text alternative to the audio in a video. Quality captions include:

• Dialogue accuracy — words spoken match the captions, including filler words like "um" only when they're meaningful
• Speaker identification — "[Dr. Smith]:" or "[Student]:" when speakers change
• Sound effects + music notation — "[applause]", "[upbeat music]", "[door slams]"
• Punctuation — sentences end with periods; questions get question marks
• Sync timing — captions appear and disappear with the audio (within ~1 second)

Free workflow — YouTube Studio (works for any video)

YouTube has the best free auto-captioning in 2026. Even if you don't want to publish to YouTube, you can use it as a captioning service by uploading videos as "unlisted":

1. Sign in to YouTube Studio.
2. Upload your video. Set visibility to Unlisted (only people with the link can see it).
3. Wait ~5 minutes. YouTube auto-generates captions in the background.
4. Once ready, click the video → Subtitles → English (auto) → Duplicate and Edit.
5. YouTube's editor shows the transcript next to the video. Play, listen, fix errors. Expect 10–15 minutes per 10 minutes of video.
6. Click Save Draft as you go, then Publish when done.
7. Now go to Subtitles → English → ⋯ → Download .srt. That's your caption file.
8. Upload the .srt file alongside your video in Canvas / Panopto / Studio.

YouTube auto-captioning is the highest-accuracy free option in 2026. It even handles speaker changes reasonably well.

Live + recorded workflow — Otter.ai

Otter joins your Zoom meetings as a participant and produces a transcript in real time. After the meeting, you get an editable transcript you can export as .srt or .vtt for Canvas.

Otter's free tier gives you 300 minutes of transcription per month. That's enough for ~10 hours of weekly lectures. Their paid tier ($16.99/mo) bumps that to 1,200 minutes.

Workflow:

1. In Otter, connect your Zoom account once.
2. When you start a Zoom meeting, Otter joins automatically and transcribes.
3. After the meeting, review the transcript in Otter — fix names, technical terms, and obvious errors.
4. Export → SRT or VTT format → upload alongside the Zoom recording.

Built-in Zoom captions

Zoom turns on automatic captions during the meeting for free (Settings → In Meeting (Advanced) → Closed captioning). After the meeting:

• Save the cloud recording. Zoom processes the transcript automatically.
• Open the recording in Zoom's web portal → click "Audio Transcript".
• Download the .vtt file — that's your caption.

Quality: ~80% accurate. Good for casual office hours, not good enough for posted lecture content without editing.

For polished, audit-grade captions — Rev.com

Rev.com uses human transcribers to deliver 99% accurate captions in 24 hours for $1.50 per video minute. For a 50-minute lecture, that's $75. For a 12-week course with weekly recorded lectures, that's $900 — meaningful, but cheap compared to the staff cost of doing it yourself.

When to use Rev: anything that will be re-used across multiple semesters, anything legally high-stakes (course content for required classes), or anything technical-vocabulary-heavy where auto-captions struggle.

Comparison table

How to add captions to a video in Canvas

Once you have your .srt or .vtt file:

1. Open the Canvas page where the video lives.
2. Click the video in the editor.
3. Look for the "Subtitles & captions" or "CC" option in the toolbar (location varies by Canvas version).
4. Upload the .srt or .vtt file.
5. Save the page. The captions are now available — students see a "CC" button on the video player.

If your video is hosted in Canvas Studio (formerly Arc), captions live in the Studio interface — go to the video → Captions tab → upload your file.

The 5-minute editing pass that matters

Auto-captions almost always butcher these:

• Proper names — your name, students' names, place names
• Technical terminology — discipline-specific vocabulary, formulas, scientific names
• Acronyms — auto-captioners often spell them out wrong or as common words
• Sentence breaks — auto-captions often run sentences together

You don't need to perfect every word. Fix the names, the technical terms, the obvious errors, and the sentence breaks — that gets you from 85% to ~95% in 5 minutes per 10 minutes of video.

References

• WCAG 2.1 — 1.2.2 Captions (Prerecorded)
• ADA.gov — Web accessibility guidance
• WebAIM: Captions, Transcripts, and Audio Descriptions