Google Videos Scraper

What does Google Videos Scraper do?

Google Videos Scraper searches public Google Videos results for videos from YouTube, TikTok, Instagram, Facebook, Vimeo, Dailymotion, publishers, and other public websites. It turns Google result cards into a stable, typed dataset suitable for research, discovery, monitoring, and API workflows.

The Actor supports one query or a batch of up to 20 queries, Google country and language localization, platform filters, publication windows, duration filters, SafeSearch, bounded pagination, stable IDs, and global deduplication. It does not log in to video platforms, solve CAPTCHAs, or crawl private engagement endpoints.

Why use Google Videos Scraper?

• Discover public videos around products, brands, people, events, and research topics.
• Monitor publisher coverage across multiple search topics.
• Restrict results to a supported platform such as YouTube or Vimeo.
• Find recently published or short, medium, and long videos.
• Schedule searches and consume datasets through the Apify API, webhooks, and integrations.
• Use the optimized HTTP-only GOOGLE_SERP transport in Apify Cloud.
• Retain a fingerprint-consistent Playwright fallback for direct, Residential, and custom proxies.

What data can be extracted?

Each dataset record includes source-backed fields that Google actually displays or that can be safely derived from the public result URL:

Field	Type	Description
id	string	Stable ID derived from a platform video ID or canonical URL
title	string	Public Google result title
url	string	Canonical public video URL
videoId	string	Platform video ID when safely derivable
platform	string	Recognized platform or source domain
sourceName	string	Source label displayed on the Google card
channelName	string	Channel or creator displayed by Google, when available
durationSeconds	integer	Parsed duration in seconds
viewCount	integer	Parsed public view count when shown
publishedAt	string	ISO-8601 time derived from the displayed date
thumbnail	string	Public thumbnail URL when exposed
isLive, isShort	boolean	Included only when the result gives explicit evidence
query, position, page	mixed	Search, ranking, and pagination context

Optional values are omitted when the Google card does not display them. The Actor does not fabricate likes, comments, followers, resolution, tags, categories, verification state, or detail-page metadata merely to fill columns.

How to scrape Google videos

1. Select Single query or Multiple queries.
2. Enter a search phrase such as machine learning tutorial.
3. Optionally select a platform, publication window, duration, country, language, and SafeSearch setting.
4. Set maxItems and maxPages, then run the Actor.
5. Open the dataset or download JSON, CSV, Excel, XML, RSS, or HTML.

Batch mode allocates records fairly across unique queries and interleaves the final result set. Limits are global, so a run cannot unexpectedly collect hundreds of items per query.

Input options

• mode, query, and queries configure single or batch search.
• platform applies a user-friendly Google site: filter.
• timeRange supports the past hour, day, week, month, or year.
• videoDuration supports any, short, medium, and long Google duration filters.
• maxItems, maxPages, and maxConcurrency keep runs bounded and fast.
• proxyConfiguration defaults to the verified GOOGLE_SERP HTTP transport. Do not tunnel Playwright through this proxy group. Direct, Residential, or custom proxy settings use browser mode.

Example input

{
  "mode": "batch",
  "query": "space documentary",
  "queries": ["healthy cooking tutorial"],
  "platform": "all",
  "timeRange": "month",
  "videoDuration": "any",
  "maxItems": 10,
  "maxPages": 2,
  "proxyConfiguration": {
    "useApifyProxy": true,
    "apifyProxyGroups": ["GOOGLE_SERP"]
  }
}

Example output

{
  "id": "google-video-52bcaebae9b413419bcb721d",
  "type": "video",
  "position": 1,
  "title": "Example public video",
  "url": "https://www.youtube.com/watch?v=example",
  "videoId": "example",
  "platform": "YouTube",
  "channelName": "Example Channel",
  "duration": "12:34",
  "durationSeconds": 754,
  "query": "space documentary",
  "country": "US",
  "language": "en",
  "scrapedAt": "2026-08-29T10:00:00.000Z"
}

The OUTPUT key-value-store record contains run counts, transport, duplicates removed, and application runtime without raw responses or proxy credentials.

Performance, cost, and reliability

The default production path fetches small HTML result pages and normally completes in seconds. Cost depends on query count, pages, proxy selection, retries, and frequency. GOOGLE_SERP and Residential traffic may have separate proxy charges. Start with 5–20 records and one or two pages.

Temporary network errors use bounded retry and progressive delay. Explicit CAPTCHA or unusual-traffic pages fail closed without repeated retries or storing challenge content. Genuine no-results searches complete successfully with an empty dataset.

FAQ and responsible use

Why are likes and comments missing?

Google result cards do not reliably expose those values. Fetching them would require separate platform APIs, logins, or additional detail crawls. This Actor preserves only public, source-backed search metadata.

Can GOOGLE_SERP be used with Playwright?

No. It is an HTTP search-result transport. The Actor automatically selects HTTP mode for that group and browser mode for other configurations.

This Actor extracts publicly displayed video-search metadata. Results can contain names or other personal data published in titles or channel names. Process personal data only with a legitimate reason and follow applicable law and target terms.

Use the Actor's Issues tab for support and its API tab for programmatic access.