Search YouTube with one API call

Read YouTube — video search with no API quota — without building or maintaining a scraper. Innertube, the same interface the site itself uses. Same response shape as every other platform here, so one parser handles all of them. 1 credit a call.

The call

curl -H "x-api-key: $SS_KEY" \
  "https://api.socialsearchapi.com/v1/youtube/search?query=llm+agents"

This call costs 1 credit once you are on credits. A new account gets 5 free calls on any platform — verify an email and this command runs as-is.

What comes back

A real response from this endpoint, captured on 2026-07-25 and trimmed to one item:

{
  "success": true,
  "platform": "youtube",
  "data": {
    "total": 3,
    "items": [
      {
        "id": "qP7LzZqGh30",
        "platform": "youtube",
        "type": "video",
        "title": "Rust Tutorial for Beginners - Full Course (With Notes and Project Ideas)",
        "url": "https://www.youtube.com/watch?v=qP7LzZqGh30",
        "author_name": "Harkirat Singh",
        "created_at": "2024-07-24T17:42:15Z",
        "metrics": {
          "views": "337,973 views"
        },
        "extra": {
          "published_raw": "2 years ago",
          "duration": "4:01:40",
          "created_at_is_approximate": true
        }
      }
    ]
  },
  "credits_used": 1
}

Items from YouTube carry: author_name, created_at, extra, id, metrics, platform, title, type, url.

Why not just scrape it yourself

The official Data API bills search at 100 quota units against a 10,000-unit daily default, so roughly 100 searches a day before you are cut off until midnight Pacific. This path has no quota to run out.

• Cost here: 1 credit a call, $0.003 at list.
• Last verified 2026-07-25: 763ms for this call.
• Coverage: Public videos. No comments, no private or unlisted videos, no watch history.

Other platforms