What is a TikTok transcript?
A TikTok transcript is the spoken audio of a TikTok video written out as text you can select, edit and reuse. TikTok does not give you one: the captions you see on screen are drawn into the video frame as pixels, so they can be read but never copied.
The distinction matters. Captions are text displayed over a video for viewers. A transcript is a standalone text document of what was said, detached from the video. TikTok ships the first and withholds the second, which is why the words of a clip you can watch in forty seconds still take twenty minutes to retype by hand.
How do you get a transcript from a TikTok video?
Paste the TikTok URL into the converter on this page and the transcript comes back as a Markdown document. Nothing is installed, no video file is saved to your machine, and you never leave the browser.
The audio is pulled from the public video and transcribed with speech recognition, which returns punctuated sentences. If the clip is not public, or if you already have the file on disk, upload the MP4, MOV or extracted audio directly instead. Both routes produce a Markdown transcript; the uploaded file also gets a timestamp on each segment.
Why can’t you copy the captions TikTok already shows?
TikTok’s auto-captions are rendered into the video image, not stored beside it as selectable text. There is no copy button, no export, and no caption file to save. Transcribing the audio again is the only way out.
| Property | TikTok’s on-screen captions | Transcript from this page |
|---|---|---|
| Selectable text | No: pixels in the frame | Yes: a .md file |
| Punctuation | Minimal, often absent | Full sentences |
| Heading | None | A Markdown heading, then the detected language |
| Timestamps | Not extractable | On an uploaded file, one per segment |
| Someone else’s public video | Read-only on screen | Same as your own |
| Ready to paste into an AI | No | Yes |
Which platforms accept a pasted link?
Paste a public link from YouTube, TikTok, Instagram, X (Twitter), Facebook, SoundCloud or Twitch. Private, restricted or removed media may be inaccessible. For a webinar recording, voice memo or client call, upload the file itself.
| Platform | Paste a link | Upload a file |
|---|---|---|
| TikTok | Yes | Yes |
| YouTube | Yes | Yes |
| Yes | Yes | |
| X (Twitter) | Yes | Yes |
| Yes | Yes | |
| SoundCloud | Yes | Yes |
| Twitch | Yes | Yes |
| Any other audio or video | No | Yes |
Why Markdown instead of a plain text transcript?
Markdown is plain text carrying a thin layer of structure: # for headings, ** for emphasis, - for lists. Language models were trained on enormous quantities of it, so they read a Markdown transcript far more reliably than an undifferentiated wall of words.
A Markdown transcript gives the model the words in full, punctuated sentences under a heading. Ask ChatGPT or Claude to rework the part after the hook for a newsletter and it can see where one idea ends and the next begins. Hand the same model a flat paste of unpunctuated caption text and it has to guess.
The output also stays useful outside the chat window: a .md file drops straight into Notion, Obsidian, a Git repository or a static site with no conversion step. Video to Markdown and Audio to Markdown produce the same kind of file, so transcripts from different sources sit side by side in one library.
How much does a TikTok transcript cost?
Video transcription is billed at 2 credits per minute of media, and a free account includes 50 credits every month. A 45-second TikTok rounds up to one minute, so the free monthly allowance covers roughly 25 short videos.
Short-form video is the cheapest thing on the site to transcribe: most TikToks run 15 to 60 seconds, against 20 to 60 minutes for the podcasts and webinars people usually put through YouTube to Markdown.
What does not work?
Private videos, friends-only videos and deleted videos cannot be fetched from a link: TikTok will not serve them to anyone, and no tool changes that. Upload the file instead when you have it.
- Music-only clips with no speech return an empty transcript. There is nothing spoken to write down.
- Text typed onto the screen is not read: the transcript comes from the audio track, not from the video frames. When the whole message is on-screen text, run a screenshot through Image to Markdown instead.
- Loud music over a quiet voice lowers accuracy. Clear voice-over with music underneath transcribes cleanly.
- Several links at once work, one per line, but this is a web tool: nothing walks a profile or a hashtag for you.
Longer background on the trade-offs: how to transcribe a TikTok video, transcribing TikToks for repurposing, and why AI can’t listen to your audio.