Niconico Scraper — Japan Video Search, Views, Comments & Tags

Search Japan's original video platform by keyword or exact tag and pull view counts, comment counts, likes, mylist saves, genres, full tag lists and upload dates — 1.9 million videos deep on a single term.

What does the Niconico Scraper do?

This Actor searches Niconico (nicovideo.jp), the video platform Japan built before YouTube arrived there and where a distinct culture still lives: comments scroll across the video, tags are edited collectively, and a song or meme's whole lineage is traceable through them. It reads Niconico's public snapshot search index, so no login and no key are involved.

Two search modes matter here. Keyword mode reads titles, descriptions and tags, and returns the loose, broad set you would expect. Exact-tag mode matches the tag as a whole — and on Niconico the tag is the unit of culture, so tag-exact search is what makes a dataset about a franchise, a producer or a fandom complete rather than approximate. Add a genre, a minimum view count or an upload-date window, choose a sort order, and the Actor pages through 100 videos per request writing one flat row each.

Who is it for?

• Music and anime rights holders tracking how their catalogue circulates on Japan's own platform.
• Trend and culture researchers studying Vocaloid, doujin music, game commentary and meme lineage.
• Marketing teams measuring a Japanese campaign's reach outside YouTube and TikTok.
• Data scientists building Japanese-language video datasets with real engagement figures.
• Journalists and academics who need evidence about Japanese internet culture rather than anecdote.

Use cases

• Pull every video tagged 初音ミク above a view threshold and rank producers by total views.
• Track how a new song's tag spreads week by week after release using the upload-date filter.
• Compare comment counts against view counts — Niconico's comment ratio is a far stronger engagement signal than likes.
• Build a genre-level view of what Japan is watching that YouTube data cannot show.
• Collect an artist's whole Niconico footprint including uploads by fans, via exact-tag mode.
• Feed an AI agent live Japanese video data so it can answer questions about current trends.

Why use this Niconico Scraper?

• Exact-tag search, not just keyword — the precise way to collect a fandom, a franchise or a producer.
• Real engagement metrics: views, comments, likes and mylist saves, all on one row.
• Millions of videos reachable — a single common term matches over 1.9 million.
• Filters that narrow properly: genre, minimum views, and upload-date windows anchored to Japan time.
• Keyless and login-free — nothing to register, nothing to rotate.
• Full tag list per video, comma-separated and ready to explode into a tag graph.

What data can you extract?

One row per video. Anything the source left blank comes back as null.

Field	Type	Description
videoId	string	Niconico content ID, e.g. sm15630734
url	string	Direct link to the video
title	string	Video title
description	string	Description with HTML stripped
viewCount	number	Total views
commentCount	number	Total comments — Niconico's strongest engagement signal
likeCount	number	Likes
mylistCount	number	Times saved to a user's mylist
lengthSeconds	number	Duration in seconds
genre	string	Niconico genre, e.g. 音楽・サウンド
categoryTags	string	Category tag such as VOCALOID
tags	string	Full tag list, comma-separated
uploaderId	number	Uploader's user ID
uploaderUrl	string	Link to the uploader's profile
channelId	number	Channel ID, for official channel uploads
channelUrl	string	Link to the channel
publishedAt	string	Upload timestamp in Japan time
lastCommentAt	string	Timestamp of the most recent comment
thumbnailUrl	string	Thumbnail image
scrapedAt	string	When this row was collected

Sample output

{
  "videoId": "sm15630734",
  "url": "https://www.nicovideo.jp/watch/sm15630734",
  "title": "『初音ミク』千本桜『オリジナル曲PV』",
  "description": "「上弦の月」～音楽劇千本桜より～ 投稿しました→sm21360174 ★★★★★★★★★★2/11 一千万再生有難う御座います！！！！！！！！！！ 様々な作品と共に千本桜が成長する事を願って 今回は桜な曲をお届けし...",
  "viewCount": 18709712,
  "commentCount": 1582813,
  "likeCount": 43512,
  "mylistCount": 333538,
  "lengthSeconds": 244,
  "genre": "音楽・サウンド",
  "categoryTags": "VOCALOID",
  "tags": "VOCALOID, 初音ミク, 黒うさP, 三重の人, 千本桜（黒うさP）, 一斗まる, ミクオリジナル曲, VOCALOID和風曲, 初音ミク名曲リンク, 重要ニコニコ文化財, VOCALOID神話入り",
  "uploaderId": 449061,
  "uploaderUrl": "https://www.nicovideo.jp/user/449061",
  "channelId": null,
  "channelUrl": null,
  "publishedAt": "2011-09-17T19:00:28+09:00",
  "lastCommentAt": "2026-08-15T02:15:42+09:00",
  "thumbnailUrl": "https://nicovideo.cdn.nimg.jp/thumbnails/15630734/15630734.34480305",
  "scrapedAt": "2026-08-15T07:46:53.516Z"
}

How to use the Niconico Scraper

Option A — a keyword sweep

Enter a term such as 音楽, ゲーム実況 or vocaloid, keep keyword mode, sort by most viewed, and set a result limit. This is the broad view.

Option B — an exact tag

Switch the mode to exact tag and enter a real Niconico tag such as 初音ミク or 東方. Only videos actually carrying that tag come back, which is what makes fandom-level datasets complete.

Option C — a release window

Set an upload-date range and sort by newest first to capture everything published in a period — useful for tracking a release, an event or a seasonal trend.

Input parameters

Input	Type	Description
query	string	Keyword or exact tag to search for. Japanese and English both work, e.g. 音楽, 初音ミク, ゲーム実況, vocaloid. Default: "音楽".
searchMode	string	Keyword searches titles, descriptions and tags. Exact tag matches the whole tag only, which is the precise way to collect a fandom or franchise. Default: "keyword".
genre	string	Optional Niconico genre filter, written in Japanese, e.g. 音楽・サウンド, ゲーム, アニメ, エンターテイメント, 動物, 料理.
minViews	integer	Optional. Skip videos below this view count — useful for filtering an archive down to what actually got watched.
publishedAfter	string	Optional date, YYYY-MM-DD. Only videos uploaded on or after this day.
publishedBefore	string	Optional date, YYYY-MM-DD. Only videos uploaded on or before this day.
sortBy	string	Order the search results before collecting them. Default: "-viewCounter".
maxResults	integer	Stop after this many videos. Each request returns 100. Default: 500.
proxyConfiguration	object	Residential proxy pinned to Japan. The snapshot index is public, but a Japanese exit is the most reliable.

Tips for best results

• Exact-tag mode needs a real tag. If a tag returns nothing, switch to keyword mode to find how the community actually spells it, then go back.
• commentCount relative to viewCount is the metric Niconico veterans read; a 1:10 ratio is extraordinary anywhere else.
• Set a minimum view count to strip the long tail out of an archive-scale query.
• Dates are anchored to Japan time (UTC+9) automatically, so a plain 2026-01-01 behaves as a Japanese calendar day.
• channelId is filled for official channel uploads — around 80% of anime results and almost none of the game-commentary ones.

Integrations

Connect this Actor to Make, Zapier, n8n, Slack, Google Sheets, GitHub, Airtable or any HTTP endpoint through Apify integrations. Every finished run can push its dataset straight into your warehouse, or fire a webhook so a downstream job starts the moment the data lands.

API usage

Run the Actor from your own code with the Apify API. Datasets can be exported as JSON, CSV, Excel, XML, RSS or HTML, and every run is available through the Apify API reference.

curl -X POST "https://api.apify.com/v2/acts/logiover~niconico-video-scraper/runs?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"input": {"query": "初音ミク", "searchMode": "tag", "minViews": 1000, "sortBy": "-viewCounter", "maxResults": 1000, "proxyConfiguration": {"useApifyProxy": true, "apifyProxyGroups": ["RESIDENTIAL"], "apifyProxyCountry": "JP"}}}'

Python, JavaScript, PHP and CLI clients are documented under Apify API clients.

Use with AI agents (MCP)

This Actor is callable from any MCP-compatible client — Claude, Cursor, VS Code or your own agent — through the Apify MCP server. An agent can call it to answer live questions about Japanese video trends, view counts and tag culture on Niconico instead of guessing from stale training data.

FAQ

Do I need a Niconico account or API key?

No. The Actor reads Niconico's public snapshot search index, which needs neither.

What is the difference between keyword and exact-tag mode?

Keyword mode searches titles, descriptions and tags and returns anything related. Exact-tag mode matches only videos carrying that exact tag. For a franchise or a fandom, exact-tag is the one that gives you a complete set.

How many videos can one run return?

A common Japanese term matches millions — 音楽 alone returns over 1.9 million. The practical limit is the result cap you set.

Can I search in English?

Yes, English terms work, but Niconico is a Japanese platform and Japanese terms return far more. vocaloid works; ボーカロイド and 初音ミク work better.

Does it download the videos?

No. It collects metadata and engagement figures, plus the thumbnail URL. No video files are downloaded.

Why is likeCount sometimes zero on very old videos?

Likes were added to Niconico long after it launched. A 2009 video can have millions of views, hundreds of thousands of comments and comparatively few likes.

Why did my run return zero videos?

In exact-tag mode, the tag probably does not exist as written. Check it in keyword mode first.

Can I filter by uploader?

Not directly, but every row carries uploaderId and uploaderUrl, so filtering the output by uploader after the run works well.

Can I export to CSV or Excel?

Yes. Every run's dataset exports as JSON, CSV, Excel, XML, RSS or HTML from the Storage tab, the API, or automatically through an integration.

How fresh is the data?

Every run reads the site live at that moment, so the data is as fresh as the site itself. Schedule the Actor hourly, daily or weekly to build a time series.

How often is the Actor updated?

It is monitored and fixed when the site changes its markup or its endpoints. Report anything that looks wrong through the Issues tab and it gets picked up.

Is it legal to scrape Niconico?

This Actor collects only publicly available information — the same pages any visitor can open without logging in. It does not bypass a login, and it does not touch private or personal accounts. Public data collection is legal in most jurisdictions, but how you use the data is your responsibility: if any record contains personal data, GDPR and comparable laws still apply, and you need a lawful basis for processing it. When in doubt, take legal advice. See Apify's ethical web scraping guide.

Related scrapers

• YouTube Search Scraper — YouTube search results and metrics.
• TikTok Hashtag Scraper — TikTok videos by hashtag.
• Twitch Scraper — live streaming data.
• MyAnimeList Scraper — anime titles, scores and rankings.
• Mercari Japan Scraper — Japanese marketplace listings.