How to Add Subtitles to a Video for TikTok

The fastest way to add subtitles to a video for TikTok is to select auto captions on the editing page after recording or uploading, then correct the words it misheard before you post. Here's the full workflow, and what auto captions can't do.

Open the free Subtitle Generator →

Add subtitles to a video for TikTok with auto captions

TikTok's auto captions automatically generate subtitles from the spoken audio. Creators select them on the editing page after they've uploaded or recorded a video, and the transcription happens without anyone typing.

The important second half: TikTok states that creators have the power to edit the text of their captions once they're generated. Do that pass every time. Speech recognition is reliable on ordinary sentences and unreliable on exactly the words that matter most. Your name, your product, the technical term the whole video is about.

TikTok's own creative guidance backs the effort up, describing captions as one of the most useful additions to a video: they provide supporting details and context, and it reports they can boost view time, brand affinity, likability, recall and uniqueness.

A caption layer is something the viewer can turn off

This is the distinction that decides which route you take. TikTok documents that viewers can switch auto captions off, by opening the share panel, tapping the captions button, and setting captions to off.

That's the right behaviour for an accessibility feature, and it's the wrong behaviour for text you consider part of the video. If your subtitles carry a joke, a correction, or the only readable version of something mumbled, a viewer toggle sitting between them and the audience is a problem.

Subtitles rendered into the picture don't have that toggle. They're pixels, so they display wherever the file plays, on every platform, forever. The trade is that they're permanent: fixing a typo means re-exporting.

Subtitles from a transcript your audio never leaves your device to produce

Montaj transcribes the audio on your device using WhisperKit. The audio track isn't sent to a server to be turned into text. From that transcript it generates word-by-word karaoke subtitles with timing at the level of the individual word, editable inline, so correcting a misheard name doesn't mean re-syncing the line by hand. The export is 1080x1920 with the captions burned into the picture and no watermark. If the reason you're adding subtitles to a video for TikTok is that people watch without sound, the words might as well be part of the image.