TikTok Transcript API: Captions, Text and SRT

To get a TikTok transcript with an API, send the video's link to GET /v1/tiktok/videos/transcript. The answer holds the whole transcript as one text, plus timed segments you can turn into subtitles. It works on any public video that has captions, with no TikTok login and no developer account.

Why TikTok's developer APIs won't do itCopy Anchor Link

Not one most developers can use. TikTok's Display API returns only the signed-in user's own videos, and its Research API, which can return speech-to-text, is limited to approved non-commercial researchers. Neither gives you the transcript of an arbitrary public video for a product.

Get a transcript in one callCopy Anchor Link

Set your key and base URL:

export API_KEY="your-api-key"
export API_URL="https://api.socialparse.dev"

Then pass the video's link (or its id, as video_id):

curl -G -H "x-api-key: $API_KEY" "$API_URL/v1/tiktok/videos/transcript" \
  --data-urlencode "url=https://www.tiktok.com/@nike/video/7587811642650545421"

Short links (vm.tiktok.com/...) aren't accepted: open one in a browser and use the full link it leads to.

What comes backCopy Anchor Link

{
  "data": {
    "video_id": "7587811642650545421",
    "language": "eng-US",
    "text": "Hey Caitlin, can you make it from there? Hey Caitlin, how about from here? ...",
    "segments": [
      { "start": 6.14, "end": 7.24, "text": "Hey Caitlin," },
      { "start": 7.82, "end": 9.16, "text": "can you make it from there?" },
      { "start": 13.82, "end": 15.88, "text": "Hey Caitlin, how about from here?" }
    ]
  },
  "page": null,
  "meta": {
    "platform": "tiktok",
    "endpoint": "video_transcript",
    "params": { "video_id": "7587811642650545421" },
    "request_id": "req_61fd34ea3abe4686",
    "api_version": "v1",
    "fetched_at": "2026-10-19T09:30:28Z",
    "cached": false
  }
}

• text is the whole transcript, ready for search, summaries or an LLM prompt.
• segments are the captions in order; start and end are seconds from the start of the video.
• language is the caption track's language, as TikTok codes it.

The transcript is shortened here. See Responses for the envelope and meta.

Choose a caption languageCopy Anchor Link

Some videos carry more than one caption track: the creator's own, auto-generated ones, and translations. The video endpoint lists them in captions, each with its language and whether it's auto generated. Pass one as language:

curl -G -H "x-api-key: $API_KEY" "$API_URL/v1/tiktok/videos/transcript" \
  --data-urlencode "video_id=7587811642650545421" \
  --data-urlencode "language=eng-US"

Leave language out to get the language spoken in the video.

Save it as an SRT subtitle fileCopy Anchor Link

The segments map straight onto SRT, the subtitle format most editors and players read:

def srt_time(seconds):
    millis = round(seconds * 1000)
    hours, millis = divmod(millis, 3_600_000)
    minutes, millis = divmod(millis, 60_000)
    secs, millis = divmod(millis, 1000)
    return f"{hours:02}:{minutes:02}:{secs:02},{millis:03}"


lines = []
number = 1
for segment in transcript["segments"]:
    lines.append(str(number))
    lines.append(f"{srt_time(segment['start'])} --> {srt_time(segment['end'])}")
    lines.append(segment["text"])
    lines.append("")
    number += 1

with open("video.srt", "w", encoding="utf-8") as file:
    file.write("\n".join(lines))

Transcripts for a whole profileCopy Anchor Link

Pair the transcript endpoint with profile videos to transcribe a creator's recent uploads, for example to research a niche or index your own brand's videos:

import os

import requests

API_URL = os.environ["API_URL"]
HEADERS = {"x-api-key": os.environ["API_KEY"]}

videos = requests.get(
    f"{API_URL}/v1/tiktok/profiles/videos", params={"handle": "nasa"}, headers=HEADERS, timeout=120
).json()["data"]

for video in videos:
    response = requests.get(
        f"{API_URL}/v1/tiktok/videos/transcript", params={"video_id": video["id"]}, headers=HEADERS, timeout=120
    )
    if not response.ok:
        print("skipped", video["url"], response.json()["error"]["code"])  # NOT_FOUND: no captions
        continue
    print(video["url"])
    print(response.json()["data"]["text"])
    print()

A profile's page holds up to 15 videos. That's one credit for the page, plus one per transcript.

Errors: no captions, private or deletedCopy Anchor Link

Errors cost no credits. See Errors for retries.

Reddit video transcripts tooCopy Anchor Link

Reddit-hosted videos work the same way: the post transcript endpoint answers with the same text and segments.

FAQCopy Anchor Link

Does this work on any TikTok video?Copy Anchor Link

On any public video that has captions, including auto-generated ones. Videos without captions answer 404 NOT_FOUND.

Do I need a TikTok account or developer approval?Copy Anchor Link

No. Social Parse reads public pages as a logged-out visitor. You only need an API key.

How much does a transcript cost?Copy Anchor Link

One credit per transcript. New accounts get 100 free credits; see pricing.

Can I translate the transcript?Copy Anchor Link

If TikTok has a translated caption track, pass its code as language. Otherwise, send text to a translation service or LLM.

What else can I get from a TikTok video?Copy Anchor Link

Views, likes, shares, saves, hashtags and its sound, from the video endpoint. See the TikTok scraper API guide for profiles and hashtags.