TikTok Transcript Endpoint: Parameters and Fields

Overview

One endpoint returns the transcript of one TikTok video. This page lists what you send and what you get back.

One GET endpoint, four query parameters.
JSON with segments, or SRT, VTT, TXT.
Names are illustrative; confirm against the live API.

Request

GET https://api.example.com/api/v1/tiktok/transcript?url=TIKTOK_URL&lang=en&format=json&ai_fallback=true
Authorization: Bearer tta_your_key

Parameter	Required	Meaning
`url`	Yes	Full TikTok video link, short link (vm., vt., tiktok.com/t/) or the 19-digit video ID.
`lang`	No	Two-letter language code. Default `en`. If that language is not available you get `language_unavailable` with the languages that are.
`format`	No	`json` (default), `srt`, `vtt` or `txt`.
`ai_fallback`	No	Default `true`: for a video with no caption track, try speech-to-text, which can return `202` and a job. `false` returns `no_transcript` instead.

Parameter names are illustrative and to be confirmed; they follow the planning spec and the owner’s scraper, which also has options for AI duration limits, moments and proxies that are not documented here.

JSON response

200 response, shortened (illustrative values)

{
  "success": true,
  "video_id": "7000000000000000001",
  "transcript": "Okay, three tips for a small garden. ...",
  "transcript_source": "captions",
  "language_code": "en",
  "available_languages": ["eng-US", "spa-ES"],
  "segments": [
    { "start": 0.5, "end": 2.4, "text": "Okay, three tips for a small garden." }
  ],
  "credits_charged": 1,
  "credits_remaining": 14999
}

Field	Meaning
`transcript`	The full text as one string.
`segments`	Timed pieces: `start`, `end` in seconds, and `text`.
`transcript_source`	`captions` for a TikTok caption track. A speech-to-text result carries a different value (the mock uses `speech_to_text`; real value to be confirmed).
`language_code`, `available_languages`	The language returned, and the tracks TikTok exposes for the video (for example `eng-US`).
`credits_charged`, `credits_remaining`	Credit bookkeeping (to be confirmed).

The owner’s scraper also returns post details (author, counts, hashtags, timestamps). Whether the public API passes them through is to be confirmed, so they are not documented here.

Other formats

srt, vtt and txt return plain text with Content-Type: text/plain. See formats.

When speech-to-text is needed

A video without captions may return HTTP 202 and a job. Poll it as described in async jobs.

Run against the local mock in tools/mock-api, which follows the assumed contract on these pages. It is not the real API, and the output is captured from that mock.