YouTube Search MCP Server — Video Search for AI Agents

Search YouTube videos by keyword across multiple regions. Extract video title, channel, views, duration, publish date, and thumbnails.

🧠 MCP Server — AI Agent Ready

This actor is a Model Context Protocol (MCP) server — AI assistants can call it directly to find YouTube videos:

{
  "searchQuery": "node.js tutorial",
  "maxResults": 10,
  "language": "en"
}

Returns structured video data: title, channel, views, duration, publish date — perfect for content research, video monitoring, and AI media pipelines.

Features

• 21 countries — US, UK, FR, DE, ES, IT, BR, JP, KR, IN, RU, CA, AU, MX, AR, SE, FI, DK, NO, NL, PL
• 16 interface languages — English, French, German, Spanish, Italian, Portuguese, Japanese, Korean, Hindi, Russian, and more
• 12 output fields — videoId, title, url, length, views, channel, channelId, published, description, thumbnail, badges, country
• CheerioCrawler — extracts ytInitialData JSON from page HTML, 256MB RAM, ~1s per query
• No proxy needed — standard User-Agent works
• No API key needed — works from the public search page

Quick Start

Search tutorials

{
  "searchQuery": "node js tutorial",
  "maxResults": 20,
  "countryCode": "US",
  "language": "en"
}

Search music in German

{
  "searchQuery": "pop musik 2026",
  "maxResults": 10,
  "countryCode": "DE",
  "language": "de"
}

Search Finnish content

{
  "searchQuery": "suomi",
  "maxResults": 5,
  "countryCode": "FI",
  "language": "fi"
}

Input Parameters

Field	Type	Default	Description
searchQuery	string	"node js tutorial"	Search term
maxResults	integer	20	Max videos to return (max 100)
countryCode	select	US	Content region (21 countries)
language	select	en	Interface language (16 languages)

Output Fields

Field	Type	Description
videoId	string	YouTube video ID
title	string	Video title
url	string	Full watch URL (youtube.com/watch?v=...)
length	string	Duration (e.g., "12:34", "1:05:20")
views	string	View count text (e.g., "2,155,030 views")
channel	string	Channel name
channelId	string	Channel URL
published	string	Relative publish time (e.g., "6 years ago")
description	string	Video description snippet
thumbnail	string	Thumbnail image URL (highest quality)
badges	string	Badges (e.g., "4K", "CC", "LIVE")
searchQuery	string	Original search term
country	string	Country code searched
scrapedAt	string	Timestamp of scrape

Use Cases

• Content research — find trending videos by topic across 21 regions
• Video SEO — analyze titles, descriptions, and thumbnails of top results
• Channel discovery — find creators and channels in specific niches
• Market analysis — compare video performance across countries
• Content aggregation — build video feeds and playlists

Pricing

$1.50 / 1,000 results — pay only for what you scrape. No monthly subscription.

Changelog

Version	Date	Changes
0.0.1	2026-07-24	Initial build — CheerioCrawler, ytInitialData extraction, 21 countries, 16 languages

Support

• Open an issue for bugs or feature requests
• Leave a review — it helps others discover the tool