How to Add Subtitles to a Video in Canva

1. Upload your MP4 or MOV to Canva and drop it onto a video design sized for your platform.
2. Open Text › Dynamic text › Captions, select the video clip, then click Generate captions.
3. Read every segment out loud against the audio and retype the words the AI misheard.
4. Drag the start and end edges of each caption block in the timeline until the text matches the speech.
5. Style one caption, apply it to all, then download an MP4 and upload it to the platform yourself.

What you need: a finished video edit, clean dialogue audio, and a Canva account in a browser or the desktop app. Mobile works, but the timeline handles are small and retiming is fiddly on a phone.

Plan on roughly ten to fifteen minutes for a one-minute talking-head clip. Most of that time goes into proofreading and retiming, not into generating. The generation itself is one click; the editing is where your subtitles stop looking automatic.

Do your trims, cuts and crops before you generate anything. Captions are anchored to timecode, so every cut you make afterwards shifts them out of place.

Canva gives you two different things that both get called "captions," and picking the wrong one costs you a re-export. Editable captions are text blocks generated from your audio that sit on the canvas and get burned into the exported video. Accessibility captions are a text track attached to a design so viewers or assistive technology can read along rather than something painted onto every frame.

According to the Canva Help Centre, Canva places AI-generated captions on a dedicated timeline layer so the text and timing can be edited without losing alignment to the video. That separate layer is why you can retype a word at any point without dragging the whole block back into position.

How to choose:

• Posting to TikTok, Reels, Shorts or a LinkedIn feed? Use editable captions and burn them in — most feed viewers watch muted with no caption toggle.
• Need text a viewer can switch off, or a file a platform ingests separately? That is the accessibility-caption or sidecar-file route, and Canva’s caption generator is not built to hand you a standalone .srt.
• Need a hard subtitle file for an archive or a client deliverable? Export the video from Canva, then produce the subtitle file in a dedicated tool — this walkthrough of burning subtitles into a video file covers the trade-offs between burned-in and sidecar formats.

For nine out of ten social posts, burned-in editable captions are what you actually want.

Time: two to five minutes, plus upload. What you need: the exported video file and a design size that matches the platform.

Start a design first, then upload. Choose a vertical 1080 × 1920 canvas for Reels, Shorts and TikTok, or 1920 × 1080 for YouTube and a website embed. Drag the clip from the Uploads panel onto the page so it fills the frame.

Canva accepts video uploads in MOV, GIF, MP4, MPEG, MKV or WEBM with a maximum file size of 1 GB per file, according to Canva’s upload formats and requirements help page. Anything larger needs compressing or splitting before it will go in.

Check your framing now, not later. If the speaker’s head is off-center or the shot needs reframing for vertical, fix it before captions exist — this guide to cropping video on Windows, Mac, iPhone and Android walks through doing it without squashing the image.

One more pre-flight check: play the audio through headphones. Background music mixed loud over dialogue produces garbled captions, and no amount of retyping is faster than lowering the music bed by a few decibels first.

Time: under a minute of clicking, plus processing. What you need: the clip already on the canvas with audible speech.

In Canva’s auto caption workflow you click "Captions" under "Dynamic Text" in the Text section, select your video, then click "Generate captions," as Canva documents on its auto caption generator page. Follow that order — selecting the clip before clicking generate is what tells Canva which audio to transcribe.

If your design has several clips on the timeline, you caption the selected one. Multi-clip edits mean repeating the process per clip, so it pays to finalize your sequence first.

Processing time scales with clip length. A thirty-second clip comes back almost immediately; a ten-minute interview takes noticeably longer, and switching browser tabs mid-generation is a common way to lose track of it.

When the text appears, resist the urge to style it. Wording and timing come first, because changing a line’s length changes how long it needs to stay on screen. Styling a caption you are about to rewrite is wasted work.

Time: five to ten minutes per minute of video. What you need: headphones and the timeline panel open.

To retime a subtitle in Canva you select the caption segment and drag its start and end edges in the timeline, as described in the Canva Help Centre article on editing and managing video captions. Those two handles are the whole timing workflow — there is no separate timecode field to type into.

Work in passes rather than fixing everything at once:

1. Wording pass. Play at normal speed and correct proper nouns, product names, acronyms and numbers. Speech recognition misses those far more often than common words.
2. Length pass. Break any segment that runs longer than about two short lines. Dense blocks get skipped by scrollers.
3. Timing pass. Set each caption to appear on the first syllable of the phrase and linger a beat after the last word.

A caption that arrives late reads as a glitch; one that arrives slightly early reads as normal. Bias toward early.

If a segment refuses to line up because the speaker pauses mid-sentence, split the sentence into two captions instead of stretching one across the silence.

Time: three to eight minutes. What you need: your brand font and a hex color for the text.

Style one segment properly, then apply that look to the rest so nothing drifts. Use a heavy sans-serif, white text, and either a solid stroke or a semi-opaque box behind it — thin fonts vanish over bright footage.

Position matters more than font choice on vertical video. Keep captions clear of the bottom quarter of the frame, where TikTok’s caption text and Instagram’s action buttons sit, and clear of the top strip where usernames appear.

Canva’s video translator can translate captions into over 100 languages, according to Canva’s video translator feature page. Translated lines are usually longer than English ones, so re-check line breaks and timing after translating rather than assuming the layout survives.

Size test: shrink your browser window until the preview is roughly the size of a phone held at arm’s length. If you cannot read the subtitle at that size, nobody scrolling will. These habits for making edits look professional apply to caption styling too — consistency beats decoration.

Time: two minutes plus render. What you need: a final preview watch-through before you click download.

Download as MP4 video at 1080p for every major feed. Canva downscales uploaded 4K video to 1080p, so the stored footage stays 1080p even when you export in 4K, according to Canva’s upload formats and requirements help page — exporting 4K from 1080p source buys you file size, not detail.

When sharing straight from Canva to Instagram, Canva documents a 100 MB MP4 size cap and a 3-second minimum, 60-second maximum video duration on its social sharing limitations help page. Direct sharing is convenient and also the easiest way to get your post silently rejected.

Download the file and upload it natively instead. Native uploads let you set the cover frame, add the platform’s own caption track and keep your file under control.

Watch the exported file once, end to end, with sound off. Burned-in captions are permanent, and a typo discovered after posting costs you the view count the post already earned. For longer uploads, this retention-focused YouTube editing workflow explains where subtitle pacing helps hold viewers.

Run this before you hit download. Each item catches a failure that is expensive to fix after export.

• Edit is locked. No trims, cuts or clip reorders after captions were generated.
• Proper nouns verified. Every name, brand, city and acronym spelled correctly, checked against your script or notes.
• Numbers checked. Prices, dates and percentages read correctly — transcription errors cluster here.
• No caption longer than two short lines. Split anything denser.
• First syllable sync. Each caption appears as the phrase starts, not after it.
• Safe zones clear. Nothing sitting under platform UI at the top or bottom of the frame.
• Contrast tested. Captions readable over the brightest frame in the video, not just the opening shot.
• Consistent styling. One font, one size, one position across all segments.
• Muted watch-through complete. The video makes sense with sound off.
• Export settings confirmed. MP4, 1080p, correct aspect ratio for the destination.

Tick every line and the post goes out clean. Skip the muted watch-through and you will find the mistake in the comments.

The most common failure is generating captions before the edit is finished. You trim four seconds from the intro, every caption after that point now fires late, and you end up dragging dozens of segment edges by hand. Fix: lock your cut first, then caption — and if you already made the cut, delete the caption layer and regenerate rather than retiming segment by segment.

Other recurring problems:

• Captions parked in the UI zone. Text hidden behind TikTok’s interface. Move the block up toward the lower-middle third.
• Music drowning dialogue. Garbled transcription. Lower the music bed, re-upload, regenerate.
• Styling before proofreading. Rewritten lines break the layout you just built. Wording first, always.
• Storage ceiling hit mid-project. Canva Free accounts get 5GB of media storage while Canva Pro accounts get 100GB, according to Canva’s upload formats and requirements help page. Clear old uploads before a long-form project rather than during one.
• Thin fonts on bright footage. White text disappears over sky, snow or a whiteboard. Add a stroke or a solid box.

One habit prevents most of these: treat captions as the final step of the edit, not a decoration you bolt on halfway through.