Subtitle a Video Automatically with AI

You shot a decent video, it's clear, and it still goes unnoticed because nobody turns the sound on. Subtitling by hand easily eats 30 to 45 minutes of work for 60 seconds of video: writing the text, splitting it up, timing every line to the speech. With a transcription AI, the same video comes back subtitled in 2 to 3 minutes, and all you have left to fix is two or three words.

This guide shows you two ways to do it: with the built-in features of your phone and the social apps, or with a specialized third-party tool. You'll also see the settings that genuinely matter (text size, position, number of words per line) and the typical mistakes automatic transcription makes: words run together, homophones, proper nouns. No technical skills required: if you can drop a file into a window, you can do this. For the text part, starting with proofreading the transcript, you can get help from the Skilzy chat before going back to the subtitling tool of your choice.

How an AI actually subtitles a video

A subtitling AI does three things: it listens to the audio track, it writes the text, then it syncs each sentence to the exact moment it's spoken.

The first step is called automatic speech transcription. The model analyzes the sound and produces text. The best known in the field is Whisper, OpenAI's open source model, trained on 680,000 hours of multilingual audio. Whisper, or something derived from it, is what runs under the hood of most tools you'll come across.

The second step is timestamping: the AI ties each word to a precise moment (00:00:04,120, for example). That's what keeps subtitles from trailing half a second behind the voice, the flaw the eye notices fastest.

The third step is styling: font, size, color, word-by-word animation. At the end of it, you get one of two formats.

• A separate subtitle file (.srt or .vtt): a plain text file you upload alongside your video on YouTube, LinkedIn or your own site. Viewers can switch them on or off, and search engines read what's in them.
• Burned-in subtitles (also called hardcoded): the text is baked into the image. You can't turn it off, but it shows up everywhere, including on TikTok, Instagram and Shorts, where playback usually starts muted.

For vertical formats, burning in is almost always the right call. For a long video on YouTube or a website, keep the .srt file: it feeds the page's SEO too.

Method 1: the built-in features of your phone and the social apps

It's free, it's already installed, and for a story or a Reel you're publishing within the minute, it's often the best effort-to-result ratio there is.

No reason to hide this option: in plenty of cases, it gets the job done.

On TikTok. In the editor, open the right-hand menu, then Captions. The app transcribes automatically and offers to let you fix each line before publishing. You can generate translations too. Styling stays limited to a handful of presets.

On Instagram. Two routes: the Captions sticker you drop on top of the video in the Reels editor, or the automatic captions option in your account settings, which turns them on by default. Editing is possible but less comfortable than on TikTok.

On YouTube. Once the video is uploaded, YouTube generates automatic subtitles in YouTube Studio, under the Subtitles tab. You can edit them line by line or upload your own .srt file, which is the better move for a video you expect to have a long life. These subtitles get indexed, which is real leverage for a faceless YouTube channel powered by AI.

On your phone itself. On iPhone, Apple's Clips app displays live captions as you speak and burns them into the video. On Android, several manufacturers ship transcription inside the gallery editor, with availability varying by model and system version.

The limits of this route are always the same: the styling is locked down, you don't walk away with a file you can reuse elsewhere, and the transcription happens inside the app you publish from. If you're posting the same video on four platforms, you do the job four times.

Method 2: specialized third-party tools

These make sense when you want a highly polished look, batch processing, or a clean subtitle file to reuse.

Here are the families of tools you'll run into, in no particular order of preference.

Pricing for these services moves around a lot: check the official page before committing, especially on annual plans.

A word on running Whisper on your own machine: it's the cheapest option per use and the most respectful of your files, since nothing leaves your computer. In exchange, you have to go through the terminal and wait longer if your machine has no dedicated graphics card. It's a good option for a high volume of sensitive content, not for subtitling a Reel between two meetings.

The settings that separate subtitles people read from subtitles people ignore

A subtitle is readable when it fits on two lines maximum, stays on screen long enough to be read, and doesn't hide behind the app's interface.

Four settings do most of the work.

Reading pace. Professional subtitling standards sit around 17 characters per second and 42 characters per line for a horizontal video. Past that, the eye gives up. AI tools tend to split either too finely or too coarsely: glance at your longest lines before exporting.

Words on screen. In a vertical format, 3 to 5 words per screen with the spoken word highlighted works better than a whole sentence. On a long 16:9 video, do the opposite: complete sentences over two lines, which are easier on the eye.

The safe zone. TikTok, Instagram and YouTube Shorts stack buttons, a handle and a caption along the edges. Keep your subtitles between 20% and 70% of the frame height, leaning toward the lower middle. A subtitle hidden behind the Subscribe button is worth nothing.

Contrast. White text with a black outline, or text on a semi-opaque bar. Avoid bright colors on a light background: it looks lovely on your calibrated screen and is unreadable in direct sunlight on a phone at 30% brightness.

These rules hold for any short-form content, including when you're producing Reels and TikToks with AI in batches.

The typical mistakes of automatic transcription, and how to fix them fast

Automatic transcription trips on four specific things: homophones, words run together, proper nouns and numbers.

Proofreading 60 seconds of subtitles takes about two minutes. It's the best investment in the whole chain, because one glaring typo displayed in big letters in the middle of the screen costs you more credibility than a sloppy edit ever will.

The checklist:

• Grammatical homophones: their/there/they're, your/you're, its/it's, to/too, then/than. Models lean on context and slip most often at the start of a sentence.
• Words run together or split apart: fast speech turns want to into wanna, a lot into allot, could have into could of. Look for words that lost an apostrophe for no reason.
• Proper nouns, brands and acronyms: your company name rarely lands on the first pass. Plenty of tools accept a custom vocabulary list — fill it in once and it serves every video you make.
• Numbers and units: twenty-nine euros ninety is correct but unreadable as a subtitle. Rewrite it as €29.90. Same logic for percentages and dates.
• Punctuation and capitalization: some models hand back one long block with no periods or commas. Add the periods at minimum — they drive how lines get split.
• Jargon and niche terms: words like LLM, n8n or Supabase often come out spelled phonetically. Fix them to the spelling your audience expects.

One trick that saves real time: proofread the text before you style it. Fixing a word in a text editor takes three seconds; fixing it once the animation is applied takes thirty.

What to remember

If you publish now and then, the subtitles built into TikTok, Instagram or YouTube Studio are enough and cost nothing. If you're producing several videos a week, go through a specialized tool where transcription, styling and export all live in the same place: the time you win back is mostly on proofreading and republishing across platforms. Either way, stick to the rule of two minutes of proofreading per minute of video, keep your subtitles out of the interface zones, and stay within two lines. The rest is comfort.

The simplest way to form your own opinion: subtitle a video you've already published, using two different methods, and compare how long each one took.