How to Add Subtitles to a Video

You finished editing a video, exported it, and then remembered the part everyone forgets until the end: subtitles. Maybe a client asked for them, maybe you noticed most people watch with the sound off, maybe you just want the thing to be usable by someone who's deaf or hard of hearing. Whatever pushed you here, the good news is that adding subtitles is far less painful than it used to be. The bad news is that there are three or four ways to do it and they're not interchangeable.

Let me walk you through how video subtitles actually work, how to make a subtitle file from your audio, and how to get that file onto the video without it looking like a ransom note.

Subtitles vs. captions vs. a burned-in overlay

People use "subtitles" and "captions" to mean the same thing in casual conversation, and for most creators the distinction barely matters. Technically, subtitles assume you can hear the audio and just need the words (often used for translation), while captions also describe sound effects and speaker changes for viewers who can't hear. If you want the full breakdown, the difference between open and closed captions is worth understanding before you commit to a method, because it changes whether viewers can turn the text off.

For now, here's the split that matters for the how-to:

• A sidecar file (like an SRT or VTT) lives next to your video and gets loaded by the player. Viewers can toggle it on or off.
• Burned-in text is baked into the pixels of the video itself. It's always visible and can never be turned off.

You'll pick one based on where the video is going. Upload platforms and websites usually want the sidecar file. Social clips that autoplay in a feed usually want the text burned in, because those players don't always support caption files.

Why the file format matters

SRT is the most widely supported subtitle format. It's a plain text file with numbered blocks, each with a start time, an end time, and a line or two of text. VTT (WebVTT) is the format built for the web, and it supports a bit more styling and positioning. If you're publishing to a website with an HTML5 video player, VTT is the native choice — the WebVTT specification is what browsers actually read. For a deeper comparison, we broke down SRT vs. VTT subtitle formats in its own guide.

The practical rule: when a platform tells you which format it wants, use that. When it doesn't, start with SRT.

Step one: get a transcript with timestamps

You can't have subtitles without knowing what's said and when. Typing that out by hand for a ten-minute video is a genuine time sink — you're pausing, rewinding, and marking timecodes every few seconds. This is the part worth automating.

The workflow I'd recommend is to run your video through a transcription tool that outputs timed captions directly. With Transkio, you can drop the video into the video-to-text tool and get a transcript back, or go straight to the video subtitle generator when a caption file is the goal. The tool listens to the audio and produces text aligned to timestamps, which is exactly the raw material a subtitle file needs.

AI-generated transcripts may contain errors — please review before relying on them.

What "review before you rely on it" looks like in practice

AI transcription is good, not perfect. Here's where it tends to slip, and what to check:

• Proper nouns and names. The model guesses at spellings it hasn't seen. Scan for anyone's name, product names, and place names.
• Homophones. "Their," "there," and "they're" can come out wrong when the audio is fast or muddy.
• Numbers and units. Prices, dates, and measurements are worth a second look.
• Overlapping speech. When two people talk at once, expect a mess. Fix those lines manually.

Read the transcript once, top to bottom, before you export it as a subtitle file. Five minutes of proofreading saves you from captions that say the wrong thing in front of your whole audience.

Exporting the caption file

Once the transcript looks right, export it in your target format. Free exports on Transkio cover TXT, SRT, VTT, which is enough for most subtitle work — SRT and VTT are the two you'll reach for. If you also need the transcript as a document for repurposing later, DOCX and JSON export come in on Pro and above, but for pure subtitling you don't need them.

Step two: add the subtitle file to your video

Now you've got an SRT or VTT file. Where it goes depends on your destination.

Uploading to a platform that supports captions

Most video platforms have a captions upload field. You publish the video, find the subtitles or CC section, and upload your file. The platform matches it to the video by the timestamps inside the file — so as long as your timings are right, it lines up automatically. This is the least destructive option because you keep the original video untouched and viewers can switch the text off.

Embedding on your own website

If you're hosting the video yourself with an HTML5 player, you attach the VTT file with a track element pointing at it. The browser reads the file and renders the captions natively. No re-encoding, no quality loss. This is why VTT exists.

A quick timing sanity check

Before you publish, play the video with the subtitles on and watch the first thirty seconds and one section from the middle. You're checking that:

1. The text appears when the words are spoken, not a beat early or late.
2. Lines don't linger so long they overlap the next one.
3. No single caption is a wall of text — two lines max is the readable ceiling.

If the timing drifts, it usually means the caption file was made from a different cut of the video than the one you're publishing. Regenerate the file from the final export.

Burning subtitles into the video

When the destination won't read a sidecar file — some social feeds, some presentation setups — you burn the text in. You load the video and the SRT into a video editor, style the text, and re-export. The upside is that the captions can never fail to show up. The downside is they're permanent: you can't fix a typo without re-rendering, and you can't offer other languages without making a separate file. Burn in only when you're confident the transcript is clean.

Subtitles in another language

If your audience doesn't share the language your video is in, subtitles are how you reach them. Transkio supports 50+ languages, and translation is available on Pro and above. The honest caveat: machine translation handles clear, straightforward speech well and struggles with idioms, slang, and heavy jargon. Treat a translated subtitle track as a strong first draft. If the video is high-stakes — a legal notice, a medical instruction, a brand campaign — have a fluent speaker read the result before it goes out.

One thing Transkio doesn't do is hand you a finished, broadcast-ready translated subtitle export with no review needed. No automated tool does, whatever the marketing says. The transcript and the timing come out automated; the polish is on you.

A realistic time estimate

For a ten-minute video with reasonably clean audio, here's roughly how the work breaks down:

• Transcription: a few minutes of processing, mostly hands-off.
• Proofreading: ten to fifteen minutes if you read carefully.
• Timing check and export: five minutes.
• Uploading or embedding: a couple of minutes.

So call it half an hour for a clean ten-minute video, most of which is you reading rather than the machine working. Longer videos scale roughly linearly, and messy audio — crosstalk, background noise, thick accents — pushes the proofreading time up. That's not a knock on the tool; noisy audio is genuinely hard, and no transcriber gets it perfect.

Putting it together

The whole thing comes down to three moves. Turn the audio into timed text. Read that text and fix what the AI guessed wrong. Then either hand the caption file to your platform or burn the text into the video, depending on where it's going.

Start with a caption file rather than burned-in text whenever you can — it keeps your options open, lets viewers toggle captions, and makes fixing a typo a two-minute edit instead of a re-render. If you're regularly captioning video, get comfortable with the video subtitle generator so the transcription step stops being the bottleneck. The rest is just careful reading.