Every TikTok transcript, as text you can search.

The TikTok Transcript API: any video URL to timestamped text

100 free credits · no card

Fast speech, slang and voiceover over music are the clips this was built for. Every response carries per-segment start times, the author and the duration, so you can quote the exact second a claim was made and anyone can check it. One credit per transcript, nothing at all when a fetch fails.

@dev · 0:47200 OK
00:00Everyone is sleeping on AI agents, and I think it is because the demos all look the same.
00:04They do not just answer questions the way a chatbot does, they actually go and do the work.
00:09So it fetches the data, cleans it up, writes the summary, and files it where it belongs.
00:16We tested this against a much bigger model last month and honestly the size barely mattered.
00:23What moved the numbers was retrieval, giving it the right source material at the right moment.
00:31Half our pipeline was video, podcasts and clips like this one, and none of it was searchable.
00:38Once we turned all of that into text with timestamps, the accuracy problem mostly went away.
00:44Retrieval won on every benchmark we ran, and it was not close.
captions · en · 8 segments1 credit

Most TikToks have no usable caption track.

A caption-only fetch returns nothing for those clips. The same call transcribes the audio instead, at the same price.

With a scraper or OSS library

One fetch path, so a markup change breaks it quietly and someone on your team owns the fix. No captions means an empty result. You rent and rotate the proxies, and you eat the cost of every failed call.

With TranscriptFetch

Three fallbacks behind one request: cache, then captions, then AI transcription of the audio. Failed requests cost nothing. TikTok changes are ours to chase, not yours.

Who it is for, and how they use it

Three ways in, depending on how you work. Same endpoint, same credit, same response shape underneath all of them.

DevelopersOne endpoint

Engineers building a product

You need TikTok text in a pipeline and you do not want to own a scraper that breaks quietly.

  1. 01Create a key in the dashboard, send it as a Bearer token.
  2. 02POST the video URL to one endpoint, same shape every time.
  3. 03Store the segments next to your other sources and move on.
Read the API reference
Research and marketingRun at volume

Analysts and marketers

You are tracking what creators and customers actually say, at more volume than anyone can watch.

  1. 01Paste a link, or drop the step into an n8n workflow.
  2. 02Send up to 50 URLs in one batch call.
  3. 03Search the text and quote the second something was said.
See batch fetching
AI and agentsNo code

Teams building with assistants

Your agent needs to read a video mid-conversation and cite it without hallucinating a timestamp.

  1. 01Add the MCP server once, no deploy and no glue code.
  2. 02Ask in plain language, the assistant calls the tool itself.
  3. 03Answers come back with the timecode attached.
Set up MCP

One request, whichever way you build.

Write code if you want to. If you would rather not, the same fetch runs from a workflow tool or straight inside an assistant.

REST

Call the endpoint

One POST with the URL. Same JSON for every TikTok you send.

Read the API reference
NO CODE

Drop it in a workflow

An official n8n node. Run it on a schedule or a trigger.

See the n8n node
MCP

Let an assistant do it

Connect once and Claude or ChatGPT fetches transcripts mid-chat.

Set up MCP
curl -X POST https://transcriptfetch.com/api/v1/transcripts/video \
  -H "Authorization: Bearer $TRANSCRIPTFETCH_KEY" \
  -d '{"video": "https://www.tiktok.com/@dev/video/713772346"}'
No code required

Let the assistant fetch the TikTok itself.

Add the MCP server once and Claude, ChatGPT or Cursor can pull a transcript mid-conversation. Same key, same credit, no glue code and nothing to deploy.

https://transcriptfetch.com/mcp

ClaudeTool: transcriptfetch
What does this TikTok say about pricing?
CALLING get_transcript(video="tiktok.com/@dev/video/72…")
At 00:41 she says pricing stayed flat because retrieval, not model size, drove the win.

Questions people ask first

TikTok Transcript API vs building the pipeline yourself

TranscriptFetch
DIY scraper + speech-to-text
Type
REST API, any language
Headless browser, proxies and an STT pipeline of your own
Setup
Get a key, make a request
Weeks of plumbing before the first transcript
No captions on the clip
Transcribed with AI speech-to-text, same endpoint
Your own transcription step to build and pay for
Blocks and proxies
Handled for you, rotating infrastructure
Your IP, blocked at scale
Markup changes
Ours to chase
Breaks quietly; someone on your team owns the fix
Timestamps
Per-segment start and duration
Depends on your STT setup
Batch
Up to 50 videos per call
Manual loops
Failed fetches
Free, billed on delivery only
You pay for the compute either way
MCP server
Yes, for Claude, Cursor and others
No

TikTok itself is no way around this: neither the Display API nor creator tools expose transcript text for arbitrary public videos, so every route to the spoken words runs through the audio.

What happens when a TikTok has no captions

TranscriptFetch
A caption-only fetch
Clip has no caption track
Transcribed with AI speech-to-text
Returns nothing, which on TikTok is most clips
Changes needed in your code
None, same video endpoint
Separate fallback logic
Extra cost
Charged only when a transcript is delivered
Premium tier, or unavailable
Blocks and rate limits
Handled by our infrastructure
Yours to solve

What the endpoint returns

Native captions when the clip has them
AI transcription when it does not, which on TikTok is most of the time
Per-segment timestamps: start and duration
Share links, vm.tiktok.com short links and permalinks all accepted
Batch endpoint for up to 50 videos per call
Plain text and segmented JSON in one response
Webhooks or polling for transcription jobs
An MCP server, plus Python and JavaScript SDKs

Guides from the blog

Start with 100 free credits

Paste a TikTok link or copy the curl, whichever suits you. No card, no sales call.