Guides
How to Add Captions to Videos for TikTok and Reels (2026 Guide)
The short answer
To add captions to videos for TikTok or Reels, transcribe the audio automatically, fix the errors, show one to four words at a time synced to your speech, and place the captions in the middle third of the frame, clear of the platform buttons. Built-in auto captions cover the basics, and styled word-by-word captions hold attention better. Shino adds styled captions with emphasis words from a single chat message.
Why captions matter on short-form video
If you are looking up how to add captions to videos, you already know why. Half your viewers are on a train, in bed next to someone sleeping, or at a desk where the sound is off. Without captions, a talking-head video is a face moving its mouth.
Mute viewers are the obvious reason. There are a few more.
- Words popping on in rhythm give the eye something to follow, and a talking-head video has very little else moving.
- Accents, fast talkers, a noisy café and words like "amortization" all get easier to follow with text on screen.
- Deaf and hard-of-hearing viewers depend on them.
- Search. TikTok, Instagram and YouTube read on-screen text and speech to figure out what a video is about. Captions that match what you say give them more to go on.
What are the ways to add captions to a video?
There are four common approaches, and I often use two of them on the same video.
1. The platform's built-in auto captions
TikTok, Instagram and YouTube Shorts can all caption your video automatically, and viewers can usually switch them on or off.
They are free and fast, and you never leave the app. The downside is that they look like everyone else's: plain styling, little control over placement, and no way to highlight a key word.
I treat them as a baseline. Even when I burn in my own captions, I leave platform captions on for accessibility.
2. Burned-in captions from an editing app
These are rendered into the video itself, in your font, color and size. It is the look on most creator videos that do numbers: big, heavy words that pop on as you say them.
You get full control, and the captions look the same on every platform. The cost is time, because you fix the transcript and style everything yourself, and viewers cannot turn them off.
3. Caption files (SRT or VTT)
A caption file is plain text with timestamps. YouTube accepts them, and they are standard for long-form, courses and webinars. They can be toggled and translated. On short-form, almost nobody turns them on, and they do nothing for how the video looks.
4. An AI editor that captions as part of the edit
Some newer tools caption as one step of a full edit, after the pauses are already cut, so the timing is right the first time. Shino works this way. Captions are one line in your brief.
How to add captions to videos step by step
The tool changes. The order stays the same.
Step 1: Cut the video before you caption it
Remove pauses, false starts and repeated takes first. Every cut after captioning shifts the timing, and you end up redoing half the captions. Our guide on how to remove silence from a video covers this step.
Step 2: Transcribe automatically
Speech recognition is very good with clean audio. Let the tool write the first draft. Typing captions by hand only makes sense for a five-second clip.
Step 3: Fix the transcript
Auto transcription trips on the same things every time, so check them:
- Your name, your product, your city and any competitor you mention. "Shopify" turning into "shop if I" happens more than you would think.
- Numbers. "Fifteen" vs "fifty" is an expensive typo in a pricing video.
- Jargon, acronyms and creator slang, which often come out garbled.
- "Um" and "uh". Take them out of the captions even if a couple stay in the audio.
Step 4: Break the text into short chunks
Short-form captions work more like a beat than like movie subtitles. Show one to four words at a time. Break at natural pauses in your speech, and never in the middle of a phrase like "New / York".
Step 5: Style for readability first
Use a heavy sans-serif. Thin and script fonts disappear on a phone. Go bigger than feels right on your laptop; if it looks slightly too big there, it is probably right on mobile. White or yellow text with a dark outline or shadow reads over almost any background.
All caps feels energetic and sentence case feels calmer. Pick one and keep it.
Step 6: Highlight emphasis words
Color or scale the words that carry the meaning: the number, the result, the surprising word. "I saved $4,000 in one month" reads very differently from the same line in plain white. One or two highlighted words per sentence is the limit. Highlight five and none of them stand out.
Step 7: Place captions inside the safe zone
TikTok and Instagram draw your username, the post caption, the like button and other controls over the video. Keep captions in the middle third of the frame, roughly at chest height, so they never sit behind the interface or over your face. Our TikTok and Reels safe zones guide has pixel numbers for a 1080x1920 frame.
Step 8: Watch it on a phone
Send the export to your phone and watch it once on mute. If you can follow the whole thing without audio, the captions work. If you catch yourself squinting at one word, fix that word.
What do good TikTok and Reels captions look like?
The talking-head captions I see holding attention tend to share a few habits:
- Words or short phrases appear in sync with speech, so the text moves with you.
- The style stays the same from video to video, which makes a feed recognizable at a glance.
- A small pop or bounce as each word appears. Subtle wins here.
- Emphasis words go in your brand color. If your brand is green, highlights are green.
- When a product card or stat pill comes on screen, the captions move or pause so the frame does not get crowded.
If you like the bold, high-energy look popularized by business creators, see our breakdown of Hormozi-style captions.
Common caption mistakes to avoid
- A full sentence on screen at once. Viewers end up reading instead of watching you.
- Captions over your face. Keep them below the chin so people can see your expression.
- Text in the bottom fifth of the screen, right where the username and post caption sit.
- Transcription errors left in. One wrong word can turn a serious point into a joke, or $15 into $50.
- A new caption style every week.
- Captions drifting out of sync, which almost always means the video was cut after captioning.
Should you use platform captions or burned-in captions?
Use both if you can. Burned-in captions give you the look, the pace and the emphasis words. Platform captions, switched on where the app allows it, serve viewers who rely on their own caption settings.
If you only do one, do burned-in styled captions. On a talking-head video they double as motion on screen, and that motion is what keeps people watching.
Quick checklist: captions for short-form video
- Cut pauses and repeated takes before captioning.
- Auto-transcribe, then fix names, numbers and jargon.
- Show one to four words at a time.
- Use a heavy sans-serif font with strong contrast.
- Highlight one or two emphasis words per sentence.
- Keep captions in the middle third, below your face.
- Stay out of the bottom and right-side interface areas.
- Use the same style on every video.
- Watch once on a phone with the sound off before posting.
How to add captions in Shino from one chat
In a timeline editor, every step above is its own chore: transcribing, fixing, splitting, styling, highlighting, then nudging captions out of the way when a graphic comes in. Shino is agent first and chat first, so captions are one sentence in what you ask for.
Upload your clip and write something like "cut the pauses, word-by-word captions, highlight the numbers in green." Shino transcribes with word timing, cuts the dead air first so the captions match the final edit, picks the emphasis words and keeps everything inside the TikTok and Reels safe zones. When a stat pill or comment callout comes on, the captions make room for it. You see stills before it renders, plus a note on what it changed.
Want bigger text, or three words per line instead of four? Say it in the same chat. There is no timeline to open. Shino learns your caption style too, so the next video starts with your font, colors and highlight habits already set. We built it to be the home for creators who talk to camera, and captions are usually the first thing people ask it for.
See how it handles talking-head videos, read how Shino works, or request early access.
