
If you are looking for a TikTok transcript API, the first thing to know is that TikTok does not offer a general-purpose one. There is no public endpoint where any developer can send a video URL and get its transcript back. What exists is a research-only API with a transcript field, and a set of third-party services that fill the gap.
This guide explains the official options, then shows a REST flow that takes a TikTok URL and returns timestamped text.
Does TikTok have an official transcript API?
Partly, and not for most developers.
- Research API. TikTok's Query Videos endpoint includes a
voice_to_textfield, described as voice-to-text and subtitles for videos that have those features turned on. The Research API is an application-based programme with its own eligibility rules and terms, so it is not something you can add to a commercial product by signing up. - Other developer products. TikTok's login, display, and content posting APIs are for working with an authorised user's own account and content. They are not designed for fetching transcripts of arbitrary public videos.
So for a product feature such as "paste a TikTok link, get the text", you need either your own speech recognition pipeline or a third-party transcript API.
Two ways a TikTok transcript gets made
Understanding the source helps you judge the output.
- Native captions. Many TikTok videos have auto-generated captions that the platform produced. When they exist, they are quick to retrieve. They are also only as good as TikTok's recognition for that clip.
- Speech recognition on the audio. When a video has no caption track, the audio has to be transcribed. This takes longer and works on any video with speech.
A good transcript API tries the first and falls back to the second, and tells you which one it used. That matters because a caption track that the creator edited is more trustworthy than a raw automatic one.
TikTok transcript API quickstart with ReelWords
ReelWords exposes a REST API that accepts a public TikTok video URL. Authentication is an x-api-key header with a key that starts with rw_, created in your account settings.
1. Create a transcript job
curl -s -X POST https://api.reelwords.ai/api/v1/transcript-jobs \
-H "x-api-key: $REELWORDS_KEY" \
-H "content-type: application/json" \
-d '{"url": "https://www.tiktok.com/@user/video/1234567890", "language": "en"}'
url is required. language is an optional BCP-47 hint such as en, es, or fr. The response is 201 Created with a job object whose status is started.
2. Poll the job
curl -s -H "x-api-key: $REELWORDS_KEY" \
https://api.reelwords.ai/api/v1/transcript-jobs/JOB_ID
status moves through started, processing, then complete or failed. A completed job looks like this, trimmed:
{
"id": "2c7d0ea0-9622-4d40-9432-715fc37a47c0",
"status": "complete",
"sourceUrl": "https://www.tiktok.com/@user/video/1234567890",
"title": "Three editing mistakes to avoid",
"durationSeconds": 42.6,
"transcriptSource": "native_captions",
"language": "en",
"segmentCount": 14,
"result": {
"segments": [
{ "start": 0.0, "end": 2.8, "text": "Here are three editing mistakes" },
{ "start": 2.8, "end": 5.1, "text": "that make your videos look amateur." }
],
"preview": "Here are three editing mistakes that make your videos look amateur...",
"formats": [
{
"id": "srt",
"label": "SRT",
"extension": "srt",
"downloadEndpoint": "/api/v1/transcript-jobs/2c7d0ea0-9622-4d40-9432-715fc37a47c0/download?format=srt"
}
]
}
}
formats lists one entry per available download: srt, vtt, txt, and txt_timestamps.
transcriptSource is native_captions when the platform's own track was used, and gemini_asr when the audio was transcribed.
3. Download a file format
curl -s -H "x-api-key: $REELWORDS_KEY" \
"https://api.reelwords.ai/api/v1/transcript-jobs/JOB_ID/download?format=srt" \
-o tiktok.srt
format accepts srt, vtt, txt, and txt_timestamps.
Status codes to handle
| Code | Meaning | What to do |
|---|---|---|
201 | Job created | Store the id and start polling |
400 | Invalid body or URL | Fix the request; do not retry as is |
401 | Missing or invalid API key | Check the x-api-key header |
402 | Plan allowance used up | Check GET /api/v1/transcript-jobs/usage |
404 | Unknown job, or it belongs to another account | Check the ID |
409 | Download requested before the job completed | Keep polling |
The API does not send webhooks, so polling is the only completion signal. Poll every few seconds with a ceiling on total wait time. The complete request and response schemas are in the API docs, and a version with retries is in our Python transcription tutorial.
What to build with TikTok transcripts
- Search and tagging. Index the text of saved or competitor videos so a team can find clips by what was said.
- Content research. Compare hooks and structures across a set of videos in text form, which is faster than rewatching them.
- Repurposing. Turn a spoken video into a caption, a post, or a script draft.
- Accessibility. Produce a corrected SRT for a video you are re-publishing elsewhere.
- Agent workflows. Give an AI assistant the transcript as context. If you would rather have the agent fetch it directly, the same operation is available as an MCP tool; see TikTok transcripts with an AI agent.
The same endpoint accepts YouTube, Instagram, X, and Facebook URLs, so one integration covers the platforms described in our social video transcription API guide.
Limits and good practice
- Public videos only. Private, deleted, age-gated, or region-restricted videos cannot be retrieved. Expect a
failedjob with afailureReason, and show the user a clear message. - Music-only clips produce little text. A video with a soundtrack and no speech will return few or no segments. Lyrics are not reliably transcribed.
- Automatic text needs review. Check names, numbers, and slang before publishing a transcript as a quote.
- Respect rights. A transcript of someone else's video is still their content. Use it for analysis and reference, and get permission before republishing.
- Cache results. A video's transcript does not change often. Store it by video ID so you do not spend allowance fetching it twice.
FAQ
Is there an official TikTok transcript API?
Not a general one. TikTok's Research API returns a voice_to_text field, but access is by application under research terms. Most products use a third-party transcript API.
Can I get a transcript for any TikTok video?
For public videos with speech, usually yes. Private, removed, or restricted videos cannot be fetched, and music-only videos have little to transcribe.
Are TikTok transcripts timestamped?
Yes. ReelWords returns segments with start and end times in seconds, and SRT and VTT files with standard cue timing.
What languages are supported?
Send a BCP-47 language hint such as es or fr when you know the spoken language. The completed job reports the language it detected or used.
How long does a TikTok transcript take?
Videos with native captions return quickly. Videos that need speech recognition take longer, depending on length. Always poll with a timeout instead of assuming a fixed duration.
Do I need to download the TikTok video first?
No. You send the page URL and the service handles retrieval. If you do need the audio itself for another purpose, see how to extract audio from a video URL.
How much does it cost?
ReelWords usage is plan-based with a free tier. Current limits are on the pricing page, and GET /api/v1/transcript-jobs/usage returns your remaining allowance.
Ready to try it? Create a key, read the ReelWords features, compare plans on pricing, or check the FAQ.