Extract, translate and analyze YouTube video transcripts with one REST call. Get the full transcript of any public YouTube video with timestamps, plain text, every available caption language, AI translations into other languages, and handy metrics (word count, reading time, speaking time, reading level). No YouTube Data API quota, no OAuth, no scraping to maintain.

👉 Get your API key on RapidAPI · Open the playground · Pricing (free plan available) · Code examples

💬 Need help or a faster response? Contact us on Telegram: @pintoflow

curl "https://youtube-captions.p.rapidapi.com/transcript?videoId=TDv56whosPQ" \
  -H "X-RapidAPI-Key: YOUR_API_KEY" \
  -H "X-RapidAPI-Host: youtube-captions.p.rapidapi.com"

1. Why this API
2. What you can build
3. Quick start
4. API reference
5. Response format and errors
6. Code examples
7. Example projects
8. Tips and best practices
9. FAQ
10. Support

• Any public video with captions. Manual captions, YouTube auto-generated captions and community captions.
• Timestamps included. Every segment has start and dur (seconds), ready to turn into SRT / VTT subtitles or clickable chapters.
• Every language at once. List the caption languages of a video, fetch one, or fetch all of them in a single call.
• Built-in translation. Translate a transcript into Spanish, French, German, Japanese, Korean, Chinese, Portuguese, Russian, Italian and more, with or without timing.
• Content metrics. Word count, character count, estimated reading and speaking time, and reading level for every transcript.
• Paste any URL. /extract/videoid turns youtube.com/watch, youtu.be, embed and mobile links into a clean video id.
• Clean JSON, consistent errors, and a free plan on RapidAPI.

Ready-to-run code for most of these lives in examples/.

1. Get an API key. Open the YouTube Captions API on RapidAPI, subscribe (a free plan is available) and copy your X-RapidAPI-Key.
2. Pick a video. Any public YouTube video with captions, e.g. https://www.youtube.com/watch?v=TDv56whosPQ → video id TDv56whosPQ.
3. Call the API with the two RapidAPI headers:

curl "https://youtube-captions.p.rapidapi.com/transcript?videoId=TDv56whosPQ" \
  -H "X-RapidAPI-Key: YOUR_API_KEY" \
  -H "X-RapidAPI-Host: youtube-captions.p.rapidapi.com"

Base URL: https://youtube-captions.p.rapidapi.com · All endpoints are GET.

Bold = required. videoId is the 11-character YouTube id (e.g. TDv56whosPQ). targetLanguage is a language code such as en, es, fr, de, ja, ko, zh, pt, ru, it.

GET /transcript?videoId=VIDEO_ID

Extract the complete transcript with timing information and helpful metrics. Use it for subtitles, accessibility, content analysis and SEO, course material and podcast transcripts.

const res = await fetch('https://youtube-captions.p.rapidapi.com/transcript?videoId=TDv56whosPQ', {
  headers: {
    'X-RapidAPI-Key': 'YOUR_API_KEY',
    'X-RapidAPI-Host': 'youtube-captions.p.rapidapi.com',
  },
});
const body = await res.json();
console.log(body.data.metrics, body.data.transcript.slice(0, 3));

Example response:

{
  "success": true,
  "message": "Transcript retrieved successfully",
  "data": {
    "videoId": "TDv56whosPQ",
    "transcript": [
      { "start": "0.080", "dur": "1.680", "text": "These are the four restrictions Marcus" },
      { "start": "1.760", "dur": "2.159", "text": "Rashford will face at Barcelona..." }
    ],
    "detected_language": "en",
    "metrics": {
      "word_count": 1200,
      "character_count": 8000,
      "estimated_reading_time": "5 min",
      "estimated_speaking_time": "8 min",
      "reading_level": "Easy"
    }
  },
  "error": null
}

GET /transcript/fulltext?videoId=VIDEO_ID

The whole transcript as one block of text plus metrics. The fastest way to send a video to an LLM or a search index.

GET /transcript2?videoId=VIDEO_ID

Same transcript data with an alternative JSON structure, for applications that need a different shape.

GET /languages?videoId=VIDEO_ID

See every caption language before requesting one. Great for language pickers.

{
  "success": true,
  "message": "Available languages retrieved successfully",
  "data": [
    { "languageCode": "en", "languageName": "English", "isAutoGenerated": false },
    { "languageCode": "es", "languageName": "Spanish", "isAutoGenerated": true },
    { "languageCode": "fr", "languageName": "French", "isAutoGenerated": true }
  ],
  "error": null
}

💡 isAutoGenerated: false means manual captions written by the creator, usually the most accurate.

GET /transcript/all?videoId=VIDEO_ID

Every available language in one request. Useful for multilingual platforms, cross-language research and transcript databases.

GET /transcript/all/text?videoId=VIDEO_ID

Same as above but clean text without timing, for reading or analysis.

GET /transcript/bylanguage?videoId=VIDEO_ID&targetLanguage=LANG_CODE

The captions YouTube itself provides in that language (native, not machine-translated). Check /languages first.

GET /transcript/translated/full?videoId=VIDEO_ID&targetLanguage=LANG_CODE

Translate the transcript into the language you need.

import requests

r = requests.get(
    "https://youtube-captions.p.rapidapi.com/transcript/translated/full",
    params={"videoId": "TDv56whosPQ", "targetLanguage": "es"},
    headers={"X-RapidAPI-Key": "YOUR_API_KEY", "X-RapidAPI-Host": "youtube-captions.p.rapidapi.com"},
)
print(r.json())

GET /transcript/translated/text?videoId=VIDEO_ID&targetLanguage=LANG_CODE

The translation with the original timing preserved: perfect for synchronized subtitles in another language.

GET /extract/videoid?videoUrl=YOUTUBE_URL

Supports every common YouTube URL format:

• https://www.youtube.com/watch?v=VIDEO_ID
• https://youtu.be/VIDEO_ID
• https://www.youtube.com/embed/VIDEO_ID
• https://m.youtube.com/watch?v=VIDEO_ID

curl "https://youtube-captions.p.rapidapi.com/extract/videoid?videoUrl=https://www.youtube.com/watch?v=TDv56whosPQ" \
  -H "X-RapidAPI-Key: YOUR_API_KEY" -H "X-RapidAPI-Host: youtube-captions.p.rapidapi.com"

{ "result_": "TDv56whosPQ" }

Successful responses use { "success": true, "message": "...", "data": ..., "error": null }. Errors use the same shape with success: false:

{
  "success": false,
  "error": {
    "code": "NO_TRANSCRIPT_AVAILABLE",
    "message": "This video doesn't have captions or transcripts available"
  }
}

async function getTranscript(videoId) {
  const res = await fetch(
    `https://youtube-captions.p.rapidapi.com/transcript?videoId=${videoId}`,
    { headers: { 'X-RapidAPI-Key': process.env.RAPIDAPI_KEY, 'X-RapidAPI-Host': 'youtube-captions.p.rapidapi.com' } }
  );
  const body = await res.json();
  if (!body.success) throw new Error(`${body.error.code}: ${body.error.message}`);
  return body.data;
}

const { transcript, metrics } = await getTranscript('TDv56whosPQ');
console.log(`${metrics.word_count} words, ${metrics.estimated_speaking_time}`);

import os, requests

def get_transcript(video_id):
    r = requests.get(
        "https://youtube-captions.p.rapidapi.com/transcript",
        params={"videoId": video_id},
        headers={"X-RapidAPI-Key": os.environ["RAPIDAPI_KEY"],
                 "X-RapidAPI-Host": "youtube-captions.p.rapidapi.com"},
        timeout=60,
    )
    body = r.json()
    if not body["success"]:
        raise RuntimeError(body["error"]["code"])
    return body["data"]

data = get_transcript("TDv56whosPQ")
print(" ".join(seg["text"] for seg in data["transcript"])[:500])

<?php
$ch = curl_init("https://youtube-captions.p.rapidapi.com/transcript?videoId=TDv56whosPQ");
curl_setopt_array($ch, [
  CURLOPT_RETURNTRANSFER => true,
  CURLOPT_HTTPHEADER => [
    "X-RapidAPI-Key: " . getenv("RAPIDAPI_KEY"),
    "X-RapidAPI-Host: youtube-captions.p.rapidapi.com",
  ],
]);
$data = json_decode(curl_exec($ch), true);
foreach (array_slice($data["data"]["transcript"], 0, 5) as $seg) {
  echo $seg["start"] . "s  " . $seg["text"] . PHP_EOL;
}

All in examples/. Set RAPIDAPI_KEY and run.

• Check /languages first when you need a specific language; use /transcript/bylanguage for native captions and /transcript/translated/* when the language isn't available.
• Prefer manual captions (isAutoGenerated: false) when accuracy matters.
• Use /transcript/fulltext for LLMs and search: fewer tokens than timed segments.
• Cache transcripts on your side. A video's captions rarely change, and each call counts against your plan.
• Keep your key on a server. Don't ship X-RapidAPI-Key in a browser extension or mobile app; use a backend like proxy-server.js.
• Handle errors by error.code, and don't retry NO_TRANSCRIPT_AVAILABLE, INVALID_VIDEO_ID or VIDEO_NOT_FOUND.
• Respect creators. Transcripts belong to their creators; follow YouTube's terms and copyright rules when you republish content.

What types of videos work? Any public YouTube video that has captions: manual captions added by the creator, YouTube's auto-generated captions, or community-contributed captions.

How accurate are the transcripts? Manual captions are very accurate (95%+). Auto-generated captions are usually 80-90% accurate for clear speech; audio quality, accents, background noise and technical terms all play a part.

Can I get transcripts from private or unlisted videos? No, only public videos are accessible.

What's the difference between native and translated transcripts? /transcript/bylanguage returns captions YouTube provides in that language. /transcript/translated/* machine-translates the original transcript into the language you ask for.

Does it work with YouTube Shorts and live stream recordings? Yes, as long as the video is public and has captions.

Do I need a YouTube Data API key or OAuth? No. You only need your RapidAPI key.

Can I use it in commercial projects? Yes: content management systems, education platforms, accessibility tools, SEO and content analysis tools, AI products and more.

What happens if a video has no captions? You get a clear NO_TRANSCRIPT_AVAILABLE error.

How current is the data? Transcripts are fetched from YouTube in real time.

💬 For questions, bug reports, feature requests or integration help, contact us on Telegram: @pintoflow. Include the endpoint and the video id you used.

1. Open the YouTube Captions API on RapidAPI and subscribe (free plan available).
2. Copy your X-RapidAPI-Key.
3. Run one of the examples or call GET /transcript?videoId=TDv56whosPQ.

If this repository helps you, give it a ⭐ so others can find it.

_{Keywords: YouTube transcript API, YouTube captions API, YouTube subtitles API, get YouTube transcript, download YouTube subtitles, YouTube to text, YouTube video to text API, YouTube SRT download, YouTube VTT, YouTube transcript JSON, YouTube transcript translation, YouTube caption extractor, YouTube transcript for ChatGPT, summarize YouTube video API, YouTube RAG, YouTube speech to text, video transcription API, auto-generated captions API, multilingual subtitles API, RapidAPI YouTube API, Node.js, Python, PHP.}