YouTube Transcript & Subtitles Scraper

Give it YouTube video, Shorts, channel or playlist links (or plain video IDs) and it returns each transcript (subtitles, captions) as JSON: full text, timed segments, language, whether it was auto-generated, every caption language the video has, and the video's title, channel, length, publish date and views. From $2.48 per 1,000 transcripts (less on higher Apify plans); you pay only for rows you get.

$2.48 per 1,000 transcripts on the Free plan, down to $1.80 on higher plans

Why this actor

• Channel and playlist input with a date range, not just single video links.
• Agent-ready output: languageMatched, availableLanguages and error codes tell an agent what it got and why.
• srt and vtt formats for subtitle work, plain text for RAG and LLM summaries.
• Videos without captions, private, age-restricted or missing videos and repeats are free.

Drop-in input

Drop-in input. Inputs written for pintostudio/youtube-transcript-scraper (videoUrl, targetLanguage), johnvc/YoutubeTranscripts (youtube_url, channel_url, output_formats) and starvibe/youtube-video-transcript (language, max_videos, start_date) work as they are. pintostudio charges $10 per 1,000 and 13.6% of its runs failed (Apify Store, 30 days to 2026-10-31); here it is from $2.48, and a video without captions is a free error row, not a failed run.

As stated in the actor’s README. The full list of accepted input names is on the Store page.

What you get

Sample row (from the README example; long text shortened, lists cut to two items):

{
  "rowType": "transcript",
  "platform": "youtube",
  "videoId": "dQw4w9WgXcQ",
  "url": "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
  "title": "Rick Astley - Never Gonna Give You Up (Official Video) (4K Remaster)",
  "channelName": "Rick Astley",
  "channelId": "UCuAXFkgsw1L7xaCfnd5JJOw",
  "channelHandle": "@RickAstleyYT",
  "channelUrl": "https://www.youtube.com/channel/UCuAXFkgsw1L7xaCfnd5JJOw",
  "durationSeconds": 213,
  "publishDate": "2009-10-24T23:57:33-07:00",
  "viewCount": 1829055386,
  "likeCount": 19491653,
  "category": "Music",
  "isLiveContent": false,
  "keywords": [
    "rick astley",
    "Never Gonna Give You Up"
  ],
  "description": "The official video for “Never Gonna Give You Up” by Rick Astley. Never: The Autobiography OUT NOW!...",
  "language": "en",
  "languageName": "English",
  "isAutoGenerated": false,
  "languageMatched": true,
  "translationStatus": "not_requested",
  "translatedFrom": null,
  "availableLanguages": [
    {
      "code": "en",
      "name": "English",
      "isAutoGenerated": false
    },
    {
      "code": "en",
      "name": "English",
      "isAutoGenerated": true
    }
  ],
  "transcript": "[♪♪♪] ♪ We're no strangers to love ♪ ♪ You know the rules and so do I ♪ ...",
  "segmentCount": 61,
  "wordCount": 487,
  "segments": [
    {
      "start": 1.36,
      "duration": 1.68,
      "text": "[♪♪♪]"
    },
    {
      "start": 18.64,
      "duration": 3.24,
      "text": "♪ We're no strangers to love ♪"
    }
  ],
  "source": null,
  "scrapedAt": "2026-10-22T08:00:00.000Z"
}

Price

Live Apify price on 2026-10-31, per 1,000 rows, by Apify plan. You pay only for rows delivered; error and summary rows marked free in the README cost nothing. Plus $0.001 per GB of memory when a run starts.

Use cases

• RAG and knowledge bases: index a channel's or a course playlist's transcripts with timestamps for citations.
• Summaries and notes: feed transcript to an LLM; wordCount tells you how many tokens to expect.
• Content and competitor research: what a channel talked about this month (publishedAfter).
• Subtitles: download srt or vtt in the language you need.
• Agents: "watch" a video for the user by reading its transcript.

Run it from code or an AI agent

Actor ID arcadium/youtube-transcripts. Set APIFY_TOKEN to your Apify API token. The input below is the actor’s prefilled example; see the guide for every field.

curl -X POST "https://api.apify.com/v2/acts/arcadium~youtube-transcripts/run-sync-get-dataset-items" \
  -H "Authorization: Bearer $APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"urls":["https://www.youtube.com/watch?v=dQw4w9WgXcQ"],"languages":[],"captionType":"any","maxVideosPerSource":10,"subtitleFormat":"none"}'

// npm install apify-client
import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: process.env.APIFY_TOKEN });
const run = await client.actor('arcadium/youtube-transcripts').call({
  "urls": [
    "https://www.youtube.com/watch?v=dQw4w9WgXcQ"
  ],
  "languages": [],
  "captionType": "any",
  "maxVideosPerSource": 10,
  "subtitleFormat": "none"
});
const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items);

# pip install apify-client
import os
from apify_client import ApifyClient

client = ApifyClient(os.environ["APIFY_TOKEN"])
run = client.actor("arcadium/youtube-transcripts").call(run_input={
    "urls": ["https://www.youtube.com/watch?v=dQw4w9WgXcQ"],
    "languages": [],
    "captionType": "any",
    "maxVideosPerSource": 10,
    "subtitleFormat": "none",
})
items = client.dataset(run["defaultDatasetId"]).list_items().items
print(items)

{
  "mcpServers": {
    "apify": {
      "url": "https://mcp.apify.com/?tools=arcadium/youtube-transcripts",
      "headers": {
        "Authorization": "Bearer <APIFY_TOKEN>"
      }
    }
  }
}

# pip install langchain-apify   (reads APIFY_API_TOKEN from the environment)
from langchain_apify import ApifyActorsTool

tool = ApifyActorsTool("arcadium/youtube-transcripts")
rows = tool.invoke({"run_input": {
    "urls": ["https://www.youtube.com/watch?v=dQw4w9WgXcQ"],
    "languages": [],
    "captionType": "any",
    "maxVideosPerSource": 10,
    "subtitleFormat": "none",
}})
# or give `tool` to an agent: create_react_agent(llm, [tool])

# pip install llama-index-readers-apify
import os
from llama_index.core import Document, VectorStoreIndex
from llama_index.readers.apify import ApifyActor

reader = ApifyActor(os.environ["APIFY_TOKEN"])
documents = reader.load_data(
    actor_id="arcadium/youtube-transcripts",
    run_input={
        "urls": ["https://www.youtube.com/watch?v=dQw4w9WgXcQ"],
        "languages": [],
        "captionType": "any",
        "maxVideosPerSource": 10,
        "subtitleFormat": "none",
    },
    dataset_mapping_function=lambda item: Document(text=str(item.get("transcript") or ""), metadata={"url": item.get("url")}),
)
index = VectorStoreIndex.from_documents(documents)

# pip install crewai crewai-tools langchain-apify   (reads APIFY_API_TOKEN from the environment)
from crewai import Agent
from crewai_tools import ApifyActorsTool

tool = ApifyActorsTool(actor_name="arcadium/youtube-transcripts")
researcher = Agent(
    role="Research analyst",
    goal="Answer with data collected by the tool",
    backstory="Uses public web data.",
    tools=[tool],
)

{
  "nodes": [
    {
      "name": "Run arcadium/youtube-transcripts",
      "type": "n8n-nodes-base.httpRequest",
      "typeVersion": 4,
      "position": [
        0,
        0
      ],
      "parameters": {
        "method": "POST",
        "url": "https://api.apify.com/v2/acts/arcadium~youtube-transcripts/run-sync-get-dataset-items",
        "authentication": "genericCredentialType",
        "genericAuthType": "httpHeaderAuth",
        "sendBody": true,
        "specifyBody": "json",
        "jsonBody": "{\"urls\":[\"https://www.youtube.com/watch?v=dQw4w9WgXcQ\"],\"languages\":[],\"captionType\":\"any\",\"maxVideosPerSource\":10,\"subtitleFormat\":\"none\"}",
        "options": {
          "timeout": 300000
        }
      }
    }
  ],
  "connections": {}
}

// Extensions > Apps Script. Store your token once: Project Settings > Script properties > APIFY_TOKEN.
function importLastRun() {
  const token = PropertiesService.getScriptProperties().getProperty('APIFY_TOKEN');
  const url = 'https://api.apify.com/v2/acts/arcadium~youtube-transcripts/runs/last/dataset/items?status=SUCCEEDED&clean=true&format=json';
  const res = UrlFetchApp.fetch(url, { headers: { Authorization: 'Bearer ' + token } });
  const rows = JSON.parse(res.getContentText());
  if (!rows.length) return;
  const keys = Object.keys(rows[0]);
  const values = rows.map((r) => keys.map((k) => (typeof r[k] === 'object' ? JSON.stringify(r[k]) : r[k])));
  const sheet = SpreadsheetApp.getActiveSheet();
  sheet.clearContents();
  sheet.getRange(1, 1, 1, keys.length).setValues([keys]);
  sheet.getRange(2, 1, values.length, keys.length).setValues(values);
}

Limitations

• Only captions YouTube already has (uploaded or auto-generated). Videos without any captions give a free no_match row; the actor does not download or transcribe audio.
• Auto-generated transcripts have no punctuation and can misspell names.
• Translation: YouTube currently refuses its machine-translated tracks to anonymous clients. With translateTo, a video that has a native track in that language returns it (translationStatus: "original", common for big channels and TED talks, which carry dozens of languages); otherwise you get the original language with translationStatus: "unavailable".
• Age-restricted and members-only videos need a signed-in account and are skipped (free unavailable row). Premieres and upcoming streams have no captions yet.
• Live streams: captions of a running live stream are not available; recorded streams work once YouTube has processed them.

FAQ

Which language do I get? The first of your languages that the video has, uploaded captions before auto-generated ones. With no languages, the original spoken language (the language of the auto-generated track); a video without an auto-generated track gives YouTube's default track, usually English when there is one. availableLanguages lists everything the video offers.

A video returned no transcript. Why? Look at its error row: no_match means the video has no captions (or none of the captionType you asked for), not_found that it is private or removed, unavailable that it is age-restricted, members-only or not published yet. These rows are free.

How fast is it? About 1 to 1.5 seconds per video, 4 videos in parallel. 100 transcripts take one to two minutes.

Something looks wrong in the output. Who do I tell? Open an issue on the actor page with the run link. We read them all.

Related actors