How to get a transcript of your own TikTok videos

You recorded a ninety-second explanation of how you price your work. It went well. Now you need it as text, for a newsletter, a client email, or the subtitles on the same video somewhere else. The obvious plan is to copy it out of TikTok, and that is where the plan ends.

What TikTok gives you to copy is the caption. The caption is what you typed on the way out the door. It is not what you said, which is the part you wanted.

What people end up doing

Three things, mostly. They retype the video while pausing it every four seconds (I have done this, and I do not recommend it). They paste the link into one of the “TikTok to text” websites. Or they give up and rewrite the thing from memory, which is how the version in the newsletter ends up different from the one on camera.

The websites deserve a word. They work by taking your link and sending it to a server that does the transcribing. For a public video that is a fair trade. For your own back catalogue it means handing a stranger’s service a few hundred videos of your voice, one link at a time, and the privacy policy of each one is a separate reading assignment.

Doing it on the phone instead

FindMyTok reaches your videos through your own signed-in TikTok session, the same way the app does, and transcribes the speech on your iPhone using Apple’s speech recognition. The audio never leaves the device. (What on-device actually means, including the parts that still touch the network.)

The steps are short:

FindMyTok's Search screen with the Spoken, On screen and Both choices and two matches FindMyTok's Search screen with the Spoken, On screen and Both choices and two matches
Search by a phrase from the video. Each hit shows the post, its date and where the words were said.
A FindMyTok result with the searched phrase highlighted, its timestamp and the Open TikTok button A FindMyTok result with the searched phrase highlighted, its timestamp and the Open TikTok button
The transcript of one video, with the phrase you searched for highlighted and the second it was said.

Which format to pick

If you are reusing the video somewhere else, the subtitle files are the useful ones. They already know when each sentence starts, which is the part that takes an hour by hand.

What the text will look like

Good, and not perfect. Names, brands and slang are where it slips: a speech engine has never heard of your cousin’s bakery, and it will spell it the way it sounds. Music under the speech costs accuracy too, and so does a windy street.

Read it once before it goes anywhere with your name on it. Fixing a transcript takes a few minutes. Writing one takes the afternoon.

Where the method stops

If the transcript is only a stepping stone, and what you really want is the video where you said something, finding a TikTok by the words you remember is the better place to start.