TollNode

YouTube data and transcript API for AI agents

TollNode gives Claude, Cursor or any AI agent live YouTube data (search, channel stats, full transcripts and comments) over API and MCP, from $0.005 per call, no API key.

Ask in plain words, like 'get the transcript of this video and summarize it', and get a clean answer in seconds.

No subscription · Sign-up optional · From $0.005 per call · Charged only when it works · No Google API key needed

Claude Code

Pull MKBHD's YouTube channel stats and his latest video

TollNode · YouTube · channel$0.015 · 1 channel · 11 s

FieldValue
Subscribers21,400,000
Total views5,900,929,964
Videos1,866
Latest videoMacbook PROLED Review: It's Not About That, 5.7M views

Real result.

Works in Claude Code, Cursor, Codex, VS Code and any MCP agent.

What your agent can pull from YouTube

Every result is structured JSON your agent can summarize, quote, filter or turn into a report or spreadsheet.

What people ask their AI to do with it

Use it as a YouTube MCP server in Claude, Cursor or Codex

Connect once. Then just ask.

Paste this into Claude Code, Cursor or Codex. It connects TollNode and gives your agent a small wallet. Add a few dollars of USDC and you're done.

terminal
npx agentcash add https://tollnode.com

Copy for your AI agent

Paste this into your agent so it knows what this node does and how to call it.

prompt
You can use TollNode's YouTube node to look up public YouTube data. Call POST https://tollnode.com/api/v1/nodes/youtube with an "action" and its fields. Actions: search (query, sortBy), channel (handle), transcript (url, language), comments (url, commentSort). Each action has its own price and is charged only when it returns a result. Over MCP, call the "youtube" tool; use list_models to see prices. Return results as tables when I ask for lists.

4 actions, one node

Send action plus the fields that action needs, to the same address. Each action has its own price, charged only when it returns a result. Leave out action and you get search.

  • search$0.042

    Find videos on a topic. 10 videos: title, URL, channel and subscriber count, views, likes, comment count, publish date, duration, description and thumbnail..

    needs: query · optional: sortBy

  • transcript$0.005

    Get the full transcript of a video. Language, full transcript text, timestamped segments, video length and the other caption languages available..

    needs: url · optional: language

  • comments$0.042

    Read the comments on a video. 20 comments: text, likes, reply count, age, and whether the creator replied or hearted it. No commenter identities..

    needs: url · optional: commentSort

  • channel$0.015

    A channel's stats and its latest 10 videos. Channel name, subscribers, total views, video count, description, joined date and location, plus its latest 10 videos with views and dates..

    needs: handle

request body: search
{"action":"search","query":"home espresso machine review"}
request body: transcript
{"action":"transcript","url":"https://www.youtube.com/watch?v=…"}

POST either body to https://tollnode.com/api/v1/nodes/youtube. From an AI assistant you don't write this yourself: ask in plain words and it picks the action.

Charged only when a call returns a result. Failed or empty calls cost $0.

YouTube data without quotas or a scraper

FAQ

Is this the official YouTube API?
No. The official YouTube Data API needs a Google Cloud project, a key and a daily quota, and it doesn't return transcripts. TollNode returns public YouTube data, including transcripts, through its own API and MCP server, priced per call.
Can I use it as a YouTube MCP server in Claude?
Yes. Add https://tollnode.com/mcp in Claude Code, Claude Desktop or Cursor, and the "youtube" tool appears with every action and its price.
Do I need a YouTube account or API key?
No. Your agent calls TollNode and gets the data back. There's no Google account, API key or quota to set up.
Does this download the video file?
No. You get structured data: search results, channel stats, transcripts and comments. There's no video or audio file download.
What if a video has no captions?
The transcript action returns nothing to transcribe and you're not charged. Try the comments or search actions instead.
How many results come back per call?
Search returns 10 videos, comments return 20, and channel returns the channel's stats plus its latest 10 videos. Call again for more.
Does this work on private or members-only videos?
No. It only reads public videos, channels and comments. Private and members-only content isn't supported.

Ready in a minute

Connect your AI
For developers: endpoint, inputs and an example response

POST https://tollnode.com/api/v1/nodes/youtube

Input

FieldTypeNotes
urlstringThe video URL, e.g. https://www.youtube.com/watch?v=aircAruvnKk (for: transcript, comments)
querystringWhat to search for, e.g. home espresso machine review (for: search)
action*enumWhat to do; sets the price. search (needs query), transcript (needs url), comments (needs url), channel (needs handle). Default search (search, transcript, comments, channel)
handlestringThe channel handle without @, e.g. NatGeo (for: channel)
sortByenumOptional, default relevance (for: search) (relevance, date, views, rating)
languagestringCaption language code, e.g. en, de, es; default en (for: transcript)
commentSortenumOptional, default TOP_COMMENTS (for: comments) (TOP_COMMENTS, NEWEST_FIRST)

Example response

{
  "results": [
    {
      "id": "4MJnPVnY3_o",
      "url": "https://www.youtube.com/watch?v=4MJnPVnY3_o",
      "date": "2026-02-07T11:00:13.000Z",
      "text": "The $3,000 espresso category hasn’t changed in years — until now.\nThe Arkel Coast brings saturated group performance,…",
      "type": "video",
      "input": "home espresso machine review",
      "likes": 2000,
      "order": 3,
      "title": "The Best Espresso Machine You've Never Heard Of",
      "duration": "00:20:15",
      "hashtags": [],
      "location": null,
      "channelId": "UCu7x43fFlwbBykF1r04T_wQ",
      "fromYTUrl": "https://www.youtube.com/results?search_query=home+espresso+machine+review&sp=EgIQAQ%3D%3D",
      "subtitles": null,
      "viewCount": 180165,
      "channelUrl": "https://www.youtube.com/@DaddyGotCoffee",
      "channelName": "Daddy Got Coffee",
      "isMonetized": null,
      "thumbnailUrl": "https://i.ytimg.com/vi/4MJnPVnY3_o/maxresdefault.jpg",
      "collaborators": null,
      "commentsCount": 288,
      "isMembersOnly": false,
      "isPaidContent": false,
      "aiVideoSummary": null,
      "translatedText": null,
      "channelUsername": "DaddyGotCoffee",
      "translatedTitle": null,
      "descriptionLinks": [
        {
          "url": "https://geni.us/ArkelCoast",
          "text": "https://geni.us/ArkelCoast"
        }
      ],
      "transcriptionUrl": null,
      "commentsTurnedOff": false,
      "aiVideoDescription": null,
      "numberOfSubscribers": 80100
    }
  ]
}

Also described in the OpenAPI file and llms.txt. Setup for every app: Agents & MCP.