Naver Scraper

Export Korean search results from Naver: titles, links, snippets, publishers, dates, and ranks.

What you get

Naver handles the majority of search in South Korea, and its results look nothing like Google's. This Actor collects its organic results as structured rows across the tabs that matter, with true on-page rank preserved.

Tab	What it returns
Integrated search	The blended page Koreans actually see
Web documents	Classic web results with sitelinks
News	Articles from Korean publishers, sortable and date-filterable
Blogs	Naver Blog posts with author and recency
Cafes	Naver Cafe community posts with the cafe name
Knowledge-iN	Q&A threads
Video	Video results with channel names

Output fields

Field	Description
query, search_type, rank, page	What was searched and where this result sat
title, url, description	The result itself
publisher	Publisher, blog author, cafe name, or channel
published_label	Naver's own recency label, such as 3시간 전 or 2026.07.02.
result_source	news, blog, cafe, kin, video, shopping, or web
is_sponsored	Whether Naver marked the placement as paid
sitelinks, thumbnail_url	Extra links and image for the result
search_url, scraped_at, source	Reproducibility fields

Example output

{"query":"맛집","search_type":"cafe","rank":1,"title":"강릉 현지인 맛집 추천 필수 코스","url":"https://cafe.naver.com/ungsangjang/871328","publisher":"부산 경남 맘스홀릭 육아 생활정보","published_label":"2주 전","result_source":"cafe","is_sponsored":false}

Input parameters

Parameter	Required	Description
queries	Yes	Korean or English search terms
searchType	No	Tab to read; defaults to integrated search
sort	No	News tab: most relevant or most recent
period	No	News tab: past day, week, month, 3 months, 6 months, or year
includeSponsored	No	Keep Naver's paid placements; off by default
maxResultsPerQuery	No	Results per term; defaults to 30
proxyConfiguration	No	Optional Apify Proxy settings if Naver starts refusing the run's IP

Use cases

• Brands entering Korea can track share of voice across news, blogs, and cafes.
• SEO teams can monitor Naver rank for a keyword set the way they already do for Google.
• PR and comms can watch Korean coverage of a company or product by recency window.
• Market researchers can mine cafe and blog discussion for consumer sentiment.

Limitations

Naver's blended results are assembled per query and per tab, so counts and ordering shift between runs; the row's rank records where a result sat in that run. Dates come as Naver's relative labels, which the Actor keeps verbatim rather than guessing an exact timestamp. The image tab is not supported because it renders through a different component tree. Naver answers most datacenter addresses, but if it starts returning HTTP 418 the run stops with that message and a Korean residential proxy is the fix. Shopping results are only reachable through the blended page, as Naver's shopping search blocks automated requests outright.

Compared to alternatives

Most Korean search Actors cover one surface, usually news, and return whatever the page happened to render. This one handles both of Naver's markup generations (the current component tree and the legacy li.bx layout the cafe tab still uses), classifies each result by its true source, strips paid placements by default, and preserves rank and page so results are comparable over time.

FAQ

Is a Naver developer account required?

No. The Actor reads the public results pages; there are no credentials to register.

Why do some rows have no description?

Naver omits snippets for some result types, such as short video and image-led cards. The field is left empty rather than filled with the page's boilerplate.

Can I get more than 400 results for one term?

Naver's own paging runs out well before that on most queries. Split the topic into several narrower terms instead.

Related Actors

• Google Trends Real-Time - Traffic & News — Thirdwatch
• Google News Scraper - Fast, Pay Per Result, No Proxy — Thirdwatch
• Substack Scraper - Posts, Authors & Search — Thirdwatch
• VK Posts Scraper - Walls, Media & Engagement — Thirdwatch

Last verified: 2026-09

More scrapers at thirdwatch.dev.