YouTube Search Scraper (bulk keywords, lightweight)

Run up to 200 keywords through YouTube search in one go and get ranked results back as a table: rank, title, video id, duration, view count, publish date and thumbnail. Sort by relevance, upload date, view count or rating. You only pay for results that are actually delivered.

Why this scraper

• Built for keyword lists, not single searches. Paste 200 keywords, get one flat table with the query and the rank on every row, ready to pivot.
• Built on a lightweight route. This actor asks YouTube's own internal API instead of loading the results page, and needs roughly 3 KB per result.
• Only delivered results are charged. A keyword without results comes back as a clear error record at no cost.
• A hard cost ceiling you control. maxResultsPerQuery is the cap: one delivered result is one charged event.
• Rank is preserved. Position 1 stays position 1, which is the whole point when you are tracking visibility for a keyword.

Who this is for

Anyone tracking what ranks on YouTube for a set of keywords: content and SEO teams checking visibility, researchers sampling a topic, and pipelines that need candidate videos before pulling details or transcripts.

Who this is not for

If you want to search YouTube once, the website is faster and free. This actor earns its place from the tenth keyword onwards, and from the moment you want the results in a spreadsheet instead of on a screen.

It returns public video metadata only. No channel subscriber lists, no comments, no personal data.

Input example

{
    "queries": [
        "web scraping tutorial",
        "n8n automation",
        "apify actor"
    ],
    "maxResultsPerQuery": 50,
    "sortBy": "relevance",
    "language": "en",
    "country": "US"
}

Output example

{
    "query": "web scraping tutorial",
    "sortBy": "relevance",
    "rank": 1,
    "videoId": "Od6M0AXpcxQ",
    "url": "https://www.youtube.com/watch?v=Od6M0AXpcxQ",
    "title": "Learn Web Scraping in 5 Minutes",
    "duration": "5:12",
    "durationSeconds": 312,
    "viewCount": 157767,
    "publishedText": "6 months ago",
    "thumbnailUrl": "https://i.ytimg.com/vi/Od6M0AXpcxQ/maxresdefault.jpg",
    "status": "ok"
}

A keyword that returns nothing produces an error record instead, and is not charged:

{
    "input": "qzxwvutsrponmlkjihgfedcba",
    "status": "error",
    "errorCode": "NO_RESULTS",
    "error": "No videos found for query 'qzxwvutsrponmlkjihgfedcba'"
}

Error codes: INVALID_QUERY, INVALID_SORT, NO_RESULTS, BLOCKED. INPUT_TRUNCATED appears once when your input is longer than this actor accepts.

Related actors from steadydata

• youtube-video-details: full metadata for the results
• youtube-channel-search: the same idea, but for channels

Pricing

Pay per event: one result-listed event per delivered result. No charge for keywords that fail or return nothing, no separate platform-usage surcharge.

FAQ

Do results differ per country? Yes, and that is deliberate. Set country to the market you care about; YouTube ranks differently per region, so a Dutch ranking check needs NL.

Which sort orders are supported? Relevance (the default), upload date, view count and rating.

Are results stable enough to track over time? Search rankings move by the hour, so treat a run as a snapshot with a timestamp, not as a fixed truth. Running the same keyword list daily is the useful pattern.

Why do I get fewer results than I asked for? YouTube stops offering more at some point, which differs per keyword. You are only charged for what was actually delivered.

Is personal data collected? No. Public video metadata only.

What happens when YouTube changes something? Internal routes shift from time to time. The actor is monitored daily and fixed fast, and while it is broken you are not charged, because only delivered results cost anything.