全能 YouTube 数据抓取工具 | 批量提取视频元数据、字幕与评论

定价

Try for free

全能 YouTube 数据抓取工具 | 批量提取视频元数据、字幕与评论

marianne-turner/youtube-scraper

自动识别 URL 或关键词，批量提取 YouTube 视频、频道、播放列表全量元数据，支持字幕、评论下载，JSON/CSV 导出，适用于市场调研与竞品分析。

免费试用

Search terms可选

Keywords to search on YouTube. Each term resolves to matching results, honouring the per-type caps and the search filters below. Ignored when Direct URLs are provided (URLs take priority).

类型: array

默认:

JSON

1[
2    {
3        "string": "python tutorial"
4    }
5]

Maximum videos per search term可选

Max regular videos to collect from EACH search term and EACH channel. Set 0 to skip regular videos.

类型: integer

默认: 10

Maximum shorts per search term可选

Max Shorts to collect from EACH search term and EACH channel. Set 0 to skip Shorts.

类型: integer

默认: 0

Maximum streams per search term可选

Max live/past streams to collect from EACH search term and EACH channel. Set 0 to skip streams.

类型: integer

默认: 0

Direct URLs可选

YouTube URLs to scrape directly: a video (watch?v= / youtu.be / shorts), a channel (/@handle, /channel/UC..., /c/, /user/), a playlist (?list=), or a search-results page. Search filters do NOT apply to direct URLs. When provided, URLs take priority and search terms are ignored.

类型: array

默认:

JSON

1[]

Download subtitles可选

Fetch each video's subtitles / transcript text. Availability + language are always reported; the transcript body may be empty when YouTube gates it behind a PO/BotGuard token.

类型: boolean

默认: false

选项：

Subtitle language可选

Preferred subtitle language codes (e.g. en, es). Empty = auto (prefer English / manually-added tracks).

类型: array

默认:

JSON

1[]

Prefer automatically generated subtitles可选

When picking a subtitle track, prefer the auto-generated (ASR) track over manually-added captions.

类型: boolean

默认: false

选项：

Subtitle format可选

Format of the subtitlesText field when Download Subtitles is on. The structured transcript array is always returned too.

类型: string

默认: srt

选项：

srtvttplaintextjson

Sorting order可选

Sort order for search terms. Ignored for direct URLs.

类型: string

默认: relevance

选项：

relevancedateviewsrating

Date filter可选

Restrict search-term results by upload date. Ignored for direct URLs.

类型: string

默认: any

选项：

anyhourtodayweekmonthyear

Video type filter可选

YouTube search 'Type' chip for search terms. Ignored for direct URLs. Note: Shorts/streams selection is driven by the per-type caps above.

类型: string

默认: any

选项：

anyvideochannelplaylistmovie

Length filter可选

Restrict search-term results by duration. Ignored for direct URLs.

类型: string

默认: any

选项：

anyshortmediumlong

Features可选

Search feature filters (ignored for direct URLs). Add any of: hd, subtitles, creativeCommons, 3d, live, purchased, 4k, 360, location, hdr, vr180.

类型: array

默认:

JSON

1[]

Scrape videos published after [date]可选

Only keep channel videos published on or after this date. Pick a date from the calendar; leave empty to disable.

类型: string

Sort by可选

Order to read a channel's videos tab. 'Latest' is exact; 'Popular' / 'Oldest' are best-effort and fall back to Latest if YouTube does not expose the sort.

类型: string

默认: latest

选项：

latestpopularoldest

Include comments可选

Fetch top-level comments for each video, including author, like/reply counts and creator-heart status.

类型: boolean

默认: false

选项：

Maximum comments per video可选

Maximum top-level comments per video when Include Comments is enabled.

类型: integer

默认: 20

Max concurrency可选

How many videos to enrich in parallel. Each video fans out several sub-requests (watch page, subtitles, comments), so very high concurrency can overload a single proxy endpoint and cause empty responses — independent of how large the rotating IP pool is. Default 8 balances speed and reliability; raise it only if your proxy sustains many concurrent connections.

类型: integer

默认: 8

Per-video timeout (seconds)可选

Abandon a single video's enrichment if it exceeds this many seconds, so one stuck video can't hold up the whole run (it is recorded as a 504 timeout). Set 0 for no limit.

类型: integer

默认: 30