How to Add Captions to a Video for Free, Without Uploading It
Most short-form video now carries captions, and for good reason: a large share of it is watched with the sound off. The popular style — big words where the one being spoken pops or lights up — needs a timestamp for every single word, which is why video apps sell captions by the minute or put them behind a subscription.
You can make them on your own device instead. The caption editor in Video Studio runs a speech recognition model inside your browser, so your video is never uploaded, there is nothing to meter, and there is no watermark on the result.
What you need before you start
- A video with clear speech. Music underneath is fine; loud music over the voice lowers accuracy.
- A reasonably recent laptop or desktop and a modern browser (Chrome, Edge, Firefox or Safari).
- A one-time download: the speech model is about 400 MB. The browser keeps it afterwards, so the second video starts straight away.
Step by step
- Open the caption editor and drop your video onto it.
- In the Captions panel, press Generate captions. The first run downloads the model; after that, transcription starts immediately.
- Choose a style. Bold pop and One word suit Reels, TikTok and Shorts; Classic subtitles suits talks, tutorials and interviews.
- Scroll the caption list and fix anything misheard — click a line, type, and the timing is kept.
- Drag the Position slider so the captions sit clear of the platform's own buttons. The safe-zone guide under the preview shows where TikTok and Reels put theirs.
- Press Export and choose captions burned in, as a subtitle track, or download an SRT from the Captions panel.
The captions are made from the edit's own sound, not the original file. If you trim, cut pauses or change speed first, the words still line up — and if you edit after captioning, captions move with the footage.
Burned in, subtitle track or SRT?
- Burned in — drawn onto the picture. Always visible, looks identical everywhere, and is what Instagram, TikTok and WhatsApp need, since they ignore subtitle tracks.
- Subtitle track — stored inside the MP4 so viewers can switch it on and off. Good for longer videos played in a proper player.
- SRT or WebVTT file — upload it separately to YouTube or LinkedIn, or bring it into Premiere or Resolve. Your own corrected captions are more accurate than a platform's automatic ones.
Languages and translation
The speech model recognises 25 European languages automatically, including English, Spanish, French, German, Italian, Portuguese, Dutch, Polish, Russian and Ukrainian — you do not have to say which one it is hearing. Captions can then be translated on your device into 20 languages, including Hindi, Arabic, Chinese, Japanese and Korean. Translation works sentence by sentence, so it reads naturally, but the word-by-word highlight cannot survive it: word order changes between languages, so translated words are spread evenly across each line.
Honest limits
- Accuracy drops with heavy accents, crosstalk, and music louder than the voice. Every line is editable.
- Languages outside the 25 it recognises — Hindi speech, for example — cannot be transcribed yet, although captions can be translated into them.
- Transcription runs on your own processor, so it is quicker on a recent machine than an old one, and long videos take a while. A progress bar shows how far it has got.
Frequently asked questions
Is my video uploaded anywhere?
No. The speech model, the caption renderer and the video encoder all run inside your browser tab. The model and fonts are served from this site, not a third-party service.
Is there a limit on minutes or videos?
No. Nothing runs on a server, so there is nothing to count.
Can I caption a video I have already edited elsewhere?
Yes — export it from your editor, drop it into Video Studio, generate captions, and either export again with them burned in or just download the SRT.