YouTube Search API for videos
Search YouTube over HTTP and receive the channel name and the video runtime alongside each result, without a Google Cloud project or a daily quota to manage.
Endpoint
Results arrive in the same JSON shape as every other engine, so existing parsing continues to work.
The link points at the watch page on youtube.com.
The snippet begins with the channel name and the duration.
The rest of the snippet is the video description.
Useful for
Tutorial discovery / Media monitoring / Content research
Runtime and channel without a quota
Set `search_service` to `youtube` and the results come from YouTube videos. The endpoint, the authentication, and the response shape are the same as every other engine, and no Google Cloud project is involved.
Runtime in the snippet
The `snippet` field follows the pattern `channel | duration | description`, and the duration is human-readable (`12:30`), so a long tutorial can be told from a short clip before fetching anything.
Channel as an attribution signal
The channel name arrives with the result, which is often what decides whether a video is worth watching.
No daily quota to budget
Requests are billed against Search1API credits rather than YouTube Data API units, so there is no daily quota to plan a workload around.
Standard Search parameters
`max_results` up to 50, plus `include_sites`, `exclude_sites`, and `time_range`.
What is a YouTube search API?
A YouTube search API is an HTTP interface that returns YouTube videos as structured results. Google publishes the YouTube Data API, which requires a Cloud project, an API key, and careful accounting against a daily quota where a single search call is charged at a much higher rate than a simple read. Search1API exposes YouTube as one of the search services on its Search endpoint, returning the video title, the watch link, and a snippet carrying the channel, the runtime, and the description, in the same JSON shape it returns for its other engines. For agents that need to point a user at a tutorial or check whether a topic has video coverage, that is usually the whole requirement.
Typical workflow
Search YouTube over HTTP and receive the channel name and the video runtime alongside each result, without a Google Cloud project or a daily quota to manage.
Send a query with `search_service` set to `youtube`.
Split `snippet` on the pipe to separate channel, runtime, and description.
Set `crawl_results` when the workflow needs the watch page rather than the listing.
One request, videos with channel and runtime
A single call searches YouTube and returns the channel and duration with each result, which is usually enough to rank candidates without opening any of them.
bash
Best for
Assistants that recommend a tutorial and need its length to judge fit.
Media monitoring that tracks video coverage of a brand or topic.
Content research that surveys how a subject is being taught.
Curation tools that build playlists from a topic query.
FAQ
Do I need a YouTube Data API key?
No. Authentication is your Search1API key. The request is a POST to https://api.search1api.com/search with `search_service` set to `youtube`.
How is this different from the YouTube Data API?
The YouTube Data API requires a Google Cloud project and charges search calls heavily against a daily quota. Search1API bills one credit per search against its own balance. The trade-off is that this returns search results rather than the full Data API surface, so it does not cover captions, analytics, or channel management.
Is the video duration included?
Yes. The runtime is the second field of the snippet, in `M:SS` or `H:MM:SS` form, so it is readable without conversion.
What does a YouTube search cost?
One credit per search, and one additional credit for each page retrieved successfully through `crawl_results`. At the base top-up rate of $1 for 1,000 credits, that is $1 per 1,000 searches. Signup includes 100 free credits with no credit card.
Can I get video transcripts?
Not from the search result itself. Set `crawl_results` to fetch the watch page, which returns the page text as markdown. Transcript extraction is not part of this endpoint.
Which other video sources are supported?
The same Search endpoint accepts `bilibili` for Chinese-language video, alongside `reddit`, `x`, `github`, general web engines, and the other Chinese-language engines, for 17 engines in total on one contract.