YouTube Search Scraper By Upload Date, Duration & Type

YouTube Search Scraper — Verified Upload Date & Duration Filters

YouTube Search Scraper runs a real YouTube search for every term you supply and returns typed video, channel or playlist rows — title, view count, duration, channel identity, live status and more — as structured JSON. Every filter (upload-date window, video length, subtitles, Creative Commons, live-now, sort order) is encoded into YouTube's own sp= search parameter and was measured against a live result set, not just wired to a form field. Each row is ready to pass to an LLM, load into a spreadsheet, or feed into a monitoring pipeline. Point it at a keyword — not a URL — and it hands back what a browser would show.

What is YouTube Search Scraper By Upload Date, Duration & Type?

YouTube Search Scraper By Upload Date, Duration & Type queries youtube.com/results for each search term you give it and returns one typed row per result — video, channel or playlist — as normalized JSON. Its defining trait is that every filter it exposes was checked against a real, unfiltered search rather than assumed to work: "This week" returned a completely different set of videos than an unfiltered search (0 of 8 IDs in common), the subtitles filter changed the result set (only 4 of 8 IDs matched), and Creative Commons narrowed it to 0 of 8 shared IDs. No YouTube account, login, cookie or API key is required — the Actor reads the same public search results page any visitor sees.

Key capabilities:

• 🎬 Video, channel and playlist result rows, each with its own field set, selected with resultKind
• 🎛️ Five real search filters — sort order, upload-date window, video length, subtitles, Creative Commons — plus a live-now flag, all encoded in YouTube's actual sp= protobuf parameter
• ↕️ Re-sorting by view count or upload date after the fact, because YouTube's own sort=date/sort=viewCount blends results back toward relevance rather than applying a strict order
• 💸 Two extraction depths: full opens each result's own page for likes, comments, tags and exact timestamps through a residential proxy, or fast reads the search page only, with no proxy required
• 📂 Optional one-request playlist enrichment for video count and owning channel, plus duplicate-ID dropping across YouTube's repeated result pages

What data can you get with YouTube Search Scraper?

resultKind decides which of four result types comes back: videos, channels, playlists, or — rarely populated by YouTube's own search — movies, which reuse the video row shape.

Result Type	Extracted Fields	Primary Use Case
Video	title, viewCount, likes, commentsCount, duration, durationSeconds, publishedAt, liveStatus, availability, categories, language, keywords, hashtags, text, channelName, channelIsVerified, numberOfSubscribers, thumbnailUrl	Content research, video SEO tracking, trend and ranking monitoring
Channel	title, channelName, channelUsername, channelUrl, channelHandleUrl, numberOfSubscribers, channelIsVerified, thumbnailUrl, text	Creator and competitor discovery, influencer shortlisting
Playlist	title, playlistVideoCount, playlistOwnerName, playlistOwnerUrl, thumbnailUrl, text	Curated-content and niche-channel mapping
Movie	Same schema as video rows	Rarely returned by YouTube search; kept for input compatibility

Verified search filters, not just declared ones

Every filter row in the schema carries an audit trail instead of a promise. Each pushed row records spParam — the exact base64 value sent to YouTube — appliedFilters, an object echoing every filter you set, and sortMode, a plain-English statement of what ordering was actually applied and why. That last field matters: requesting orderResultsBy = date at detailLevel = fast returns YouTube's blended order, because fast results carry no exact publish timestamp to re-sort by, and sortMode says so explicitly rather than silently returning the wrong order. orderResultsBy = rating is passed through to YouTube but is never re-sorted, because YouTube exposes no rating figure anywhere in the response payload.

{
  "resultKind": "video",
  "searchRank": 3,
  "spParam": "EgQIAxAB",
  "appliedFilters": {
    "sortBy": "relevance",
    "uploadDate": "thisWeek",
    "duration": null,
    "resultKind": "video",
    "onlyWithSubtitles": false,
    "onlyCreativeCommons": false,
    "onlyLiveNow": false
  },
  "sortMode": "relevance (YouTube's native ordering, unchanged)",
  "searchUrl": "https://www.youtube.com/results?search_query=cold+brew+review&sp=EgQIAxAB"
}

Channel rows

Search with resultKind = channel and each row carries the channel's subscriber count, verification badge, handle (channelUsername) and both its canonical URL (channelUrl) and the raw uploader URL as YouTube returns it (channelHandleUrl). Useful for building a creator shortlist from a topic search without opening a single channel page by hand.

How does YouTube Search Scraper differ from the official YouTube Data API?

YouTube publishes an official search endpoint, search.list in the YouTube Data API v3, and it returns fewer fields per call than this Actor does at detailLevel = full.

Feature	YouTube Data API v3 (search.list)	YouTube Search Scraper
Setup	Google Cloud project and API key required	Only an Apify account; no Google credentials
Results per call	maxResults 0–50, default 5; more pages need a nextPageToken and draw further quota	resultsPerTerm 1–2000 in a single run
Result types per call	video, channel and playlist together by default, or one via type	One resultKind per run, each with its own full row schema
Channel/playlist detail	Snippet only — subscriber count and playlist size need a separate channels/playlists call	numberOfSubscribers and playlistVideoCount/playlistOwnerName returned in the same run
Per-video detail	No likes, comment count or exact duration in the search snippet — needs a further videos.list call	likes, commentsCount and durationSeconds returned directly at detailLevel = full
Video length filter	videoDuration: any/short/medium/long, enum only	Same three bands, with measured real ranges: Short 66–144 s, Medium 247–1129 s, Long 3606–57468 s
Upload window	publishedAfter/publishedBefore accept exact RFC 3339 timestamps	Preset windows only — lastHour/today/thisWeek/thisMonth/thisYear

(YouTube Data API v3 parameters as published at developers.google.com/youtube/v3/docs/search/list, checked 26 July 2026.)

The official API is the right choice when you need exact-timestamp date filtering or already manage a Google Cloud project and its quota. YouTube Search Scraper is the faster path when you want one Apify token, up to 2000 results per term, and channel or playlist detail without stitching together three separate API resources.

🧭 How to scrape YouTube with YouTube Search Scraper

1. Open YouTube Search Scraper on the Apify Console and click Try for free
2. Enter one or more terms into searchTerms — the field that drives the run; nothing else is required
3. Set the real query controls that apply: orderResultsBy, publishedWithin, videoLength, resultKind, and the onlyWithSubtitles / onlyCreativeCommons / onlyLiveNow toggles
4. Click Start
5. Download the dataset as JSON, CSV, Excel or open it through the Apify API

How to run multiple search terms in one job

searchTerms is an array — one term per line runs as one independent search, each recording which term produced every row in its input field. Terms are fetched with bounded internal concurrency (up to three in flight at once); this is not a configurable input, just how the run is scheduled internally.

⬇️ Input

Every parameter below is optional — required is empty in the schema — and every filter is encoded into the real sp= parameter described above.

Parameter	Required	Type	Description	Example Value
searchTerms	No	array	One search term per line, exactly as typed into YouTube's search box. Default ["python programming"]. Legacy key queries still accepted.	["cold brew review"]
resultsPerTerm	No	integer	Results to collect per term. Minimum 1, maximum 2000, default 10. At detailLevel = full each row costs roughly 587 KB of RESIDENTIAL proxy traffic. Legacy key maxResults still accepted.	20
orderResultsBy	No	string	Ordering to request: relevance (default), date, viewCount, rating. viewCount is re-sorted by us into strict descending order; date is re-sorted only at detailLevel = full; rating cannot be verified. Legacy key sortBy still accepted.	viewCount
publishedWithin	No	string	Upload window: "" (default, any time), lastHour, today, thisWeek, thisMonth, thisYear. Legacy key uploadDate still accepted.	thisWeek
videoLength	No	string	Video length band: "" (default, any), short (under 4 min), medium (4–20 min), long (over 20 min). Video results only. Legacy key duration still accepted.	medium
resultKind	No	string	Result type: video (default), channel, playlist, movie. Legacy key type still accepted.	video
onlyWithSubtitles	No	boolean	Restrict to videos carrying closed captions. Default false. Video results only.	true
onlyCreativeCommons	No	boolean	Restrict to Creative Commons–licensed videos. Default false. Video results only.	false
onlyLiveNow	No	boolean	Restrict to streams currently broadcasting. Default false. Video results only.	false
detailLevel	No	string	full (default) opens each result's own page for likes, comments, tags, timestamp, category and language, and requires a RESIDENTIAL proxy. fast reads the search page only, works without a proxy, at roughly 1/277th the traffic.	fast
enrichPlaylistRows	No	boolean	Adds one extra ~2 KB request per playlist result to fill in playlistVideoCount and the owning channel. Default true. Ignored when not searching for playlists.	true
dropDuplicateIds	No	boolean	Drops repeated IDs — YouTube repeats results across deep result pages (5 duplicates measured in 426 rows). Default true.	true
proxyConfiguration	No	object	Apify Proxy configuration. Empty by default (useApifyProxy: false). detailLevel = full is bot-walled on datacenter IPs and automatically escalates to RESIDENTIAL if blocked.	{"useApifyProxy": true, "apifyProxyGroups": ["RESIDENTIAL"]}
verboseLogs	No	boolean	Logs every underlying search call, the exact sp= filter used, and each proxy step. Default false. Legacy key debug still accepted.	false

Example input

{
  "searchTerms": ["cold brew review", "espresso machine comparison"],
  "resultsPerTerm": 25,
  "orderResultsBy": "viewCount",
  "publishedWithin": "thisMonth",
  "videoLength": "medium",
  "resultKind": "video",
  "onlyWithSubtitles": true,
  "detailLevel": "fast",
  "dropDuplicateIds": true
}

Common pitfall: an empty input still runs a search

searchTerms carries a schema default of ["python programming"], not just a form prefill. If you start a run through the Apify API with an empty input object ({}) rather than the Console form, the platform fills in that default before the run starts — the Actor genuinely searches "python programming" and you are billed row_result events for a query you never typed. If your integration always intends to supply its own terms, pass searchTerms explicitly on every call rather than relying on an empty body to fail safely; sending "searchTerms": [] does fail safely — the run pushes a single uncharged error row and exits. Separately, videoLength, onlyWithSubtitles, onlyCreativeCommons and onlyLiveNow only affect video results — set with resultKind = channel or playlist, they are silently ignored and the reason is written to the run log.

⬆️ Output

Every result is one dataset row of typed, normalized JSON with a consistent key set for its resultKind. Export as JSON, CSV, Excel or an RSS/XML feed, or read the dataset directly through the Apify API. Real search results are charged as row_result events; rows pushed with "type": "error" (no terms supplied, or a search that returned nothing) are not charged.

Scraped results

[
  {
    "type": "video",
    "resultKind": "video",
    "title": "Cold Brew Coffee: Full Review & Taste Test",
    "id": "dQw4w9WgXcQ",
    "url": "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
    "input": "cold brew review",
    "date": "2026-06-30T14:00:00Z",
    "publishedAt": "2026-06-30T14:00:00Z",
    "publishedAtEpoch": 1782914400,
    "text": "A full taste test of five cold brew concentrates, ranked by bitterness and cost per cup.",
    "thumbnailUrl": "https://i.ytimg.com/vi/dQw4w9WgXcQ/maxresdefault.jpg",
    "hashtags": ["#coldbrew", "#coffee"],
    "keywords": ["cold brew", "coffee review", "taste test"],
    "channelName": "Brew Lab",
    "channelId": "UCabc1234567890defghijk",
    "channelUrl": "https://www.youtube.com/channel/UCabc1234567890defghijk",
    "channelUsername": "@brewlab",
    "channelIsVerified": false,
    "numberOfSubscribers": 84200,
    "viewCount": 152340,
    "likes": 6210,
    "commentsCount": 340,
    "commentsTurnedOff": false,
    "duration": "12:47",
    "durationSeconds": 767,
    "liveStatus": "not_live",
    "availability": "public",
    "isMembersOnly": false,
    "categories": ["Howto & Style"],
    "language": "en",
    "ageLimit": 0,
    "searchRank": 3,
    "order": 0,
    "spParam": "EgQIAxAB",
    "appliedFilters": {
      "sortBy": "viewCount",
      "uploadDate": "thisMonth",
      "duration": "medium",
      "resultKind": "video",
      "onlyWithSubtitles": true,
      "onlyCreativeCommons": false,
      "onlyLiveNow": false
    },
    "sortMode": "viewCount (post-sorted by us, 25/25 rows had a view count)",
    "detailLevel": "fast",
    "searchUrl": "https://www.youtube.com/results?search_query=cold+brew+review&sp=EgQIAxAB",
    "scrapedAt": "2026-07-26T09:12:04Z",
    "runId": "aBcD1234EfGh5678"
  },
  {
    "type": "channel",
    "resultKind": "channel",
    "title": "Brew Lab",
    "id": "UCabc1234567890defghijk",
    "url": "https://www.youtube.com/channel/UCabc1234567890defghijk",
    "input": "cold brew review",
    "date": null,
    "text": "Coffee gear reviews and brewing science, weekly.",
    "thumbnailUrl": "https://yt3.ggpht.com/abc123=s176-c-k-c0x00ffffff-no-rj",
    "hashtags": [],
    "keywords": [],
    "channelName": "Brew Lab",
    "channelId": "UCabc1234567890defghijk",
    "channelUrl": "https://www.youtube.com/channel/UCabc1234567890defghijk",
    "channelUsername": "@brewlab",
    "channelHandleUrl": "https://www.youtube.com/@brewlab",
    "channelIsVerified": false,
    "numberOfSubscribers": 84200,
    "viewCount": null,
    "likes": null,
    "commentsCount": null,
    "commentsTurnedOff": null,
    "duration": null,
    "isMembersOnly": null,
    "durationSeconds": null,
    "publishedAt": null,
    "publishedAtEpoch": null,
    "liveStatus": null,
    "availability": null,
    "categories": null,
    "language": null,
    "ageLimit": null,
    "searchRank": 1,
    "order": 1,
    "spParam": "EgQIAxAB",
    "appliedFilters": {
      "sortBy": "viewCount",
      "uploadDate": "thisMonth",
      "duration": "medium",
      "resultKind": "video",
      "onlyWithSubtitles": true,
      "onlyCreativeCommons": false,
      "onlyLiveNow": false
    },
    "sortMode": "relevance (YouTube's native ordering, unchanged)",
    "detailLevel": "fast",
    "searchUrl": "https://www.youtube.com/results?search_query=cold+brew+review&sp=EgQIAxAB",
    "scrapedAt": "2026-07-26T09:12:04Z",
    "runId": "aBcD1234EfGh5678"
  },
  {
    "type": "playlist",
    "resultKind": "playlist",
    "title": "Cold Brew Reviews Playlist",
    "id": "PLxyz9876543210abcdef",
    "url": "https://www.youtube.com/playlist?list=PLxyz9876543210abcdef",
    "input": "cold brew review",
    "date": null,
    "text": "Every cold brew review from the channel, newest first.",
    "thumbnailUrl": "https://i.ytimg.com/vi/dQw4w9WgXcQ/hqdefault.jpg",
    "hashtags": [],
    "keywords": [],
    "channelName": "Brew Lab",
    "channelId": "UCabc1234567890defghijk",
    "channelUrl": "https://www.youtube.com/channel/UCabc1234567890defghijk",
    "channelUsername": null,
    "numberOfSubscribers": null,
    "viewCount": null,
    "likes": null,
    "commentsCount": null,
    "commentsTurnedOff": null,
    "duration": null,
    "isMembersOnly": null,
    "durationSeconds": null,
    "publishedAt": null,
    "publishedAtEpoch": null,
    "channelIsVerified": null,
    "liveStatus": null,
    "availability": null,
    "categories": null,
    "language": null,
    "ageLimit": null,
    "playlistVideoCount": 14,
    "playlistOwnerName": "Brew Lab",
    "playlistOwnerChannelId": "UCabc1234567890defghijk",
    "playlistOwnerUrl": "https://www.youtube.com/channel/UCabc1234567890defghijk",
    "searchRank": 2,
    "order": 2,
    "spParam": "EgQIAxAB",
    "appliedFilters": {
      "sortBy": "viewCount",
      "uploadDate": "thisMonth",
      "duration": "medium",
      "resultKind": "video",
      "onlyWithSubtitles": true,
      "onlyCreativeCommons": false,
      "onlyLiveNow": false
    },
    "sortMode": "relevance (YouTube's native ordering, unchanged)",
    "detailLevel": "fast",
    "searchUrl": "https://www.youtube.com/results?search_query=cold+brew+review&sp=EgQIAxAB",
    "scrapedAt": "2026-07-26T09:12:04Z",
    "runId": "aBcD1234EfGh5678"
  }
]

Every output field

Field	Present on	Description
type, resultKind	all	Row kind: video, channel or playlist. type mirrors resultKind for compatibility with earlier runs.
title	all	Result title
id	all	Video, channel or playlist ID
url	all	Canonical URL for the result
input	all	The search term that produced this row
date, publishedAt	videos	Exact publish timestamp, ISO 8601
publishedAtEpoch	videos	Publish timestamp as Unix seconds
text	all	Description (video/channel/playlist)
thumbnailUrl	all	Highest-resolution thumbnail offered
hashtags	videos	Hashtags parsed out of the description
keywords	videos	Cleaned tag list (internal yt:-prefixed tags removed)
channelName, channelId, channelUrl, channelUsername	all (channel identity on the row's own channel; on playlists, filled by enrichment)	Owning or represented channel
channelHandleUrl	channels	Raw uploader URL as returned by the underlying extractor
channelIsVerified	videos, channels	Verification badge
numberOfSubscribers	channels always; videos when the data is present	Subscriber count
viewCount	videos	View count
likes	videos, at detailLevel = full	Like count
commentsCount	videos, at detailLevel = full	Comment count
commentsTurnedOff	videos, at detailLevel = full	true/false; null at fast depth, meaning "not fetched", not "disabled"
duration	videos	Formatted length, h:mm:ss or m:ss
durationSeconds	videos	Length in raw seconds
isMembersOnly	videos	Derived from availability
liveStatus	videos	e.g. is_live, not_live, was_live
availability	videos	e.g. public, subscriber_only, premium_only
categories	videos	YouTube category list
language	videos	Detected/declared language
ageLimit	videos	Age restriction, 0 if none
playlistVideoCount	playlists, with enrichPlaylistRows on	Number of videos in the playlist
playlistOwnerName, playlistOwnerChannelId, playlistOwnerUrl	playlists, with enrichPlaylistRows on	Owning channel of the playlist
searchRank	all	YouTube's own position in the result page
order	all	Row position after dedup and re-sorting
spParam	all	The exact sp= value sent to YouTube for this search, or null
appliedFilters	all	Object echoing sortBy, uploadDate, duration, resultKind, onlyWithSubtitles, onlyCreativeCommons, onlyLiveNow as applied
sortMode	all	Plain-language statement of the ordering actually applied
detailLevel	all	Extraction depth actually used for this row — channel and playlist results are always extracted flat regardless of the detailLevel you set, since fully resolving a channel or playlist would mean walking every video inside it
searchUrl	all	Full YouTube search URL fetched, including the sp= filter
scrapedAt	all	UTC timestamp the row was scraped
runId	all, when available	Apify run ID
error	error rows only	Failure reason; present only on uncharged rows

How can I use the data extracted with YouTube Search Scraper?

• 📊 Video SEO and content teams pull searchRank, viewCount and publishedAt across a keyword set to see which videos currently rank and how fast newer uploads are gaining views, informing which topics and formats to target next.
• 🎥 Rights and production teams set onlyCreativeCommons = true to source reusable footage, then use channelName, channelUrl and text for licensing attribution and outreach.
• 🤖 AI engineers and LLM developers issue a query, receive typed video/channel/playlist JSON, and pass it straight into an agent's context — no HTML parsing step between the search and the model.
• 🔬 Market and competitive researchers run a channel search per topic and compare numberOfSubscribers and channelIsVerified across results to map who currently owns a niche.

How do you monitor YouTube search rankings over time?

Rank monitoring is running the same search repeatedly and watching which video IDs move, appear, or disappear from the results — trustworthy here specifically because spParam and sortMode prove the same filter was applied on every run rather than YouTube silently changing what it served you. Diff two runs of the same searchTerms on id, searchRank and order to see who gained or lost position; diff on viewCount and numberOfSubscribers to see how fast a leading result is growing; watch for id values with no prior match to catch new uploads entering the result set.

A concrete loop: schedule a run across your keyword list, keep the previous run's dataset, and after each new run match rows by id within the same input term. Alert when a video's searchRank moves into your tracked range for the first time, when viewCount crosses a threshold between runs, or when a competitor's id appears where it wasn't before. Set orderResultsBy = viewCount for growth tracking specifically — it is the one ordering the Actor verifiably re-sorts into strict descending order, so the diff reflects real movement rather than YouTube's blended relevance.

Schedule the recurring runs with Apify's own Schedule feature from the Actor's page, and use a webhook on run completion to push each dataset into your own pipeline the moment it finishes.

Integrate YouTube Search Scraper and automate your workflow

YouTube Search Scraper works with any language or tool that can send an HTTP request through the Apify API.

REST API with Python

from apify_client import ApifyClient

client = ApifyClient("<YOUR_APIFY_TOKEN>")

run = client.actor("<YOUR_USERNAME>/youtube-search-scraper-by-upload-date-duration-type").call(run_input={
    "searchTerms": ["cold brew review"],
    "resultsPerTerm": 25,
    "orderResultsBy": "viewCount",
    "detailLevel": "fast",
})

for row in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(row["title"], row["viewCount"], row["channelName"])

Scheduled monitoring and delivery

Set up a recurring run with Apify's Schedule feature on the Actor's page, choosing how often it re-runs your searchTerms list. Attach a webhook to the run to notify your own endpoint, or push the dataset onward, the moment each scheduled run finishes — no polling loop required on your end.

Is it legal to scrape YouTube search results?

Yes — YouTube Search Scraper collects only what any anonymous visitor sees on YouTube's public search results page, without logging in or accessing any private data. Courts have repeatedly held that scraping publicly accessible web data does not violate the U.S. Computer Fraud and Abuse Act (hiQ Labs, Inc. v. LinkedIn Corp., 9th Cir., 2019). Video titles, view counts and channel handles are business and public-content data rather than personal data in the GDPR/CCPA sense, though a channel or video description may incidentally contain a person's name — treat any such fields with care in downstream use. You remain bound by YouTube's Terms of Service regardless of whether the CFAA applies. Consult your legal team before using extracted data for bulk redistribution or commercial resale.

❓ Frequently asked questions

Does YouTube Search Scraper support region or language filtering?

No. Unlike the official YouTube Data API's regionCode and relevanceLanguage parameters, this Actor has no region or language input — results reflect whatever YouTube serves the request's proxy exit location. If you need consistent regional targeting, route the run through a specific Apify Proxy country in proxyConfiguration.

How does YouTube Search Scraper decide whether to return videos, channels or playlists?

The resultKind parameter — video (default), channel, playlist or movie — selects one result type per run. It is encoded into the sp= filter as YouTube's own type field, so the underlying search itself, not just post-filtering, is restricted to that kind.

How does YouTube Search Scraper handle YouTube's anti-bot measures?

detailLevel = full is bot-walled on datacenter IPs — measured directly as zero rows returned with a "Sign in to confirm you're not a bot" page — so a Full run with no proxy selected starts on a RESIDENTIAL proxy automatically, and a Full run on your own proxy escalates to RESIDENTIAL if YouTube blocks it. Every escalation is written to the run log. detailLevel = fast works on the Actor's own container IP without any proxy.

Does YouTube Search Scraper really verify that its filters work?

Yes. Each filter's real effect on the result set was measured by ID overlap against an unfiltered search: "This week" shared 0 of 8 IDs, subtitles shared 4 of 8, Creative Commons shared 0 of 8, and live-now returned 8 of 8 rows with liveStatus = is_live. That audit trail is also returned per row in spParam, appliedFilters and sortMode, so you can confirm on your own data that a filter was actually applied.

How many results does YouTube Search Scraper return per search term?

Up to resultsPerTerm, from 1 to 2000, defaulting to 10. At detailLevel = full each row costs roughly 587 KB of RESIDENTIAL proxy traffic, so 2000 results is close to 1.2 GB for a single term; the same 2000 rows at detailLevel = fast cost roughly 2 KB each. dropDuplicateIds (default true) removes repeats YouTube serves across deep result pages before this count is reached.

How do I monitor YouTube search rankings for a keyword over time?

Schedule recurring runs of the same searchTerms with Apify's Schedule feature, then diff each new dataset against the previous one by id within the same input term — watch searchRank, order and viewCount for movement, and treat any unmatched id as a new entrant to the result set.

Does YouTube Search Scraper work with Claude, ChatGPT and other AI agent frameworks?

Yes. It is callable as a standard HTTP endpoint through the Apify API, so any agent framework that can issue a request — LangChain, CrewAI, a custom tool definition — can call it and receive typed JSON rows directly, letting an agent ground its answer in a live YouTube search rather than training data.

How does YouTube Search Scraper compare to other YouTube scrapers on Apify?

Checked on the Apify Store on 26 July 2026: api-ninja/youtube-search-scraper documents a broader features toggle object (HD, 3D, 4K, 360, HDR, VR180) and a "scrape all results" mode, but does not document that its filters were verified against real result sets. grow_media/youtube-search-api wraps the official YouTube Data API directly and documents "up to 1000 video results," but is scoped to video results only and states no channel or playlist search. tcz/youtube-search-term-channel-crawler is a minimal tool focused on pulling channel contact emails from search results and does not document result filtering at all. YouTube Search Scraper's difference is the combination of typed video/channel/playlist rows, real upload-date and duration filters with measured proof they work, and a per-row audit trail (spParam, sortMode) that lets you confirm it on your own runs.

Can I use YouTube Search Scraper without managing proxies or a YouTube/Google API key?

Yes for the API key — no Google Cloud project or YouTube API key is ever required. For proxies: detailLevel = fast needs none at all. detailLevel = full requires a RESIDENTIAL proxy, and if you leave proxyConfiguration empty the Actor selects Apify's RESIDENTIAL proxy for you automatically; you never have to source or rotate one yourself.

What happens if I call YouTube Search Scraper with an empty input?

It still runs. searchTerms carries a schema default of ["python programming"], so a run started through the Apify API with {} as input is not rejected — the platform fills in that default before execution, the Actor searches "python programming," and you are charged row_result events for it. To guarantee a run only ever searches terms you explicitly chose, always include searchTerms in the input you send.

💬 Your feedback

Found a bug, or a field YouTube exposes that isn't in the output yet? Open an issue on the Actor's Issues tab on Apify — reports that include the search term and the filters you used are the fastest to reproduce and fix.