Naver Blog & Cafe Scraper

Every Naver Blog and Naver Cafe post about your keyword, with the full text, images, likes, views and comments - no API key, no daily quota and no 1,000-result ceiling.

Naver Blog and Naver Cafe are where Koreans actually write about products, restaurants, travel, parenting and brands - it is the review layer that Google does not see. The usual way in is the official Naver Search API, and it has three problems for research work: you need a developer key, you get 25,000 calls a day, and one keyword never returns more than 1,000 results. This Actor reads the same public pages a browser does, splits a date range into single days so each day gets its own 1,000, and opens every post for its full text instead of a two-line snippet. Point it at a keyword, a whole blog or a whole cafe, and get one clean row per post.

✅ What you get / ❌ what this isn't

✅ What you get	❌ What this isn't
More than 1,000 posts per keyword - a date range is searched one day at a time	Not capped at the 1,000 results the official Naver Search API returns per keyword
The full post text, every image URL and the author's tags	Not the 2-line search snippet the official API gives you
commentCount on every row, plus likeCount on blog posts and viewCount on cafe posts	Not a list of links you have to open yourself to see engagement
Comments (author, text, date, reply flag) over plain HTTP	Not a 2 GB headless-browser mode just to read comments
Naver Blog and Naver Cafe in one identical row shape	Not two different schemas you have to merge
Members-only cafe posts that Naver search links to, read through the search link's own access token	Not a login - private cafes and posts that search does not surface are out of reach
No API key, no Naver developer account, no quota to manage	Not a wrapper that stops when a shared key runs out for the day

🔎 Why use this Actor

• Day-by-day date windows. Naver answers every search with at most 1,000 posts. With dateFrom / dateTo set, each day in the range is its own search, so a busy keyword over a week returns thousands of posts, not the first 1,000. Measured on 캠핑 (camping) for 1-3 September 2026: 2,500 unique posts, every one inside the range, where a single search stops at 1,000.
• Full text, not snippets. Each post is opened and its body is flattened into readable text with paragraph breaks kept, plus images and tags. That is what sentiment analysis, topic modelling and LLM summarisation need.
• Engagement on every row. Blog rows carry the like count and comment count; cafe rows carry views and comment count too. Sort by them to find the posts that actually got read.
• Comments without a browser. includeComments reads each post's comment thread from the same JSON endpoints Naver's own pages use, so it stays fast and cheap.
• Search-link access tokens. Naver's cafe search attaches a token to each result link that opens the post for a visitor. The Actor passes it along, which recovers members-only posts that search itself shows to the public.
• Whole blogs and whole cafes. Paste blog.naver.com/<id> for a blogger's latest posts with like and comment counts, or cafe.naver.com/<cafe> for a community's latest public articles.
• Pay per post delivered. No start fee, and members-only articles that cannot be read are skipped and never charged.

👥 Who it's for

Brand, marketing and research teams who need what Korean consumers are saying, and developers feeding Korean text into models.

• Monitor every blog and cafe mention of your brand or product, day by day.
• Collect Korean product reviews at volume for sentiment or aspect analysis.
• Find the bloggers who write about your category, ranked by likes and comments, for influencer outreach.
• Track what a parenting, real-estate or hobby cafe is discussing this week.
• Build a Korean-language text corpus for LLM fine-tuning or RAG.

Example tasks

Each one is a ready-made configuration you can open, adjust and run.

• Scrape Naver Blog camping gear reviews
• Scrape Naver Blog restaurant posts in Seoul
• Monitor Korean cosmetics reviews on Naver
• Track brand mentions on Naver Blog and Cafe
• Scrape Naver Cafe parenting community posts
• Scrape Naver Cafe real estate discussions
• Collect iPhone buyer reviews from Naver
• Scrape all posts of one Naver blogger

⚙️ How to scrape Naver Blog and Cafe posts

1. Put one or more keywords in Search keywords. Korean keywords return far more posts, for example 캠핑용품 후기 rather than camping gear review.
2. Choose Where to search: Naver Blog, Naver Cafe, or both.
3. Set From date and To date for the period you care about. Every day in between is searched separately.
4. Set Max posts per keyword or URL. Leave Full post text and images on, and turn on Include comments if you need them.
5. Click Start, then open the Posts or Engagement view on the Output tab and export to CSV, Excel or JSON.

Get yesterday's posts every morning

Save your input as a Task with From date and To date left empty and Sort by set to Newest first, then schedule it daily. Without a date range the Actor walks back from today one day at a time and stops at your post limit, so a daily run always starts with the newest posts.

📥 Input

{
  "searchKeywords": ["캠핑용품 후기"],
  "source": "both",
  "sortBy": "date",
  "dateFrom": "2026-09-01",
  "dateTo": "2026-09-07",
  "maxPostsPerKeyword": 500,
  "includeContent": true,
  "includeComments": true,
  "maxCommentsPerPost": 50
}

• searchKeywords - keywords to search, one per line. Korean works best.
• source - blog, cafe or both. With both, the post limit applies to each source separately. Default blog.
• sortBy - date (newest first, default) or relevance (Naver's own ranking).
• dateFrom / dateTo - optional range in Korean time, YYYY-MM-DD. Each day is its own search, which is how a keyword goes past 1,000 posts. Only dateFrom means "up to today".
• maxPostsPerKeyword - stop after this many posts per keyword and source, or per blog or cafe URL. 0 means no limit. Default 50.
• startUrls - blog posts, whole blogs, cafe articles or whole cafes. Mobile m. links work too.
• includeContent - open each post for full text, images, tags and likes. Default on. Off gives a faster list of titles and snippets.
• includeComments - collect comments for each post. Default off.
• maxCommentsPerPost - cap per post, 0 for all. Default 50.

📤 Output

source	title	publishedAt	authorName	likeCount	commentCount	viewCount
blog	"다시는 안 간다"던 나주곰탕 노안집, 결국 또 가봤습니다	2026-09-11T16:27:05+09:00	리뷰하는 남자	44	8
cafe	횽님들 캠핑 에어컨 어떤거 쓰시나요?	2026-07-27T16:05:32+09:00	서울II은채아빠		52	1320
cafe	멀고 멀었던 광양 배알도 야영장 후기 (부자캠핑)	2026-07-12T18:34:34+09:00	준영왕자		46	2294

Every row also carries url and postId, the full text and a snippet, images (every image URL in the post), tags, authorId, blogName or cafeName with cafeUrl and cafeMemberCount, boardName (the cafe board or blog category), comments as an array of {author, text, publishedAt, isReply, likeCount} when requested, the keyword that found the post, and scrapedAt. All dates are ISO timestamps in Korean time (+09:00).

💵 How much does it cost?

$1.50 per 1,000 posts delivered, and no start fee. Comments are an extra $0.50 per 1,000 posts that have comments, charged only when you turn them on - posts with no comments are not charged for them. A week of blog posts for one busy keyword at 2,000 posts costs $3, or at most $4 with comments. Members-only cafe articles that cannot be read are skipped and cost nothing.

🔁 Run it on the Apify platform

Schedule it daily or weekly, call it from the Apify API or any of the official clients, export to CSV, JSON, Excel or XML, and connect the output to Make, Zapier, n8n, Google Sheets, a webhook or your own database.

⚠️ Limits and caveats

• 1,000 posts per day per keyword on Naver Blog. That is Naver's ceiling for one search, and a single day is the smallest window it offers. A keyword that gets more than 1,000 blog posts a day will be sampled, not complete, on its busiest days.
• Naver Cafe search is shallower. It returns roughly 100 posts per day per keyword (measured on 캠핑: 85 to 108 per day). Use several narrower keywords, or paste the cafe's own URL, when you need a community's full history.
• Private cafes stay private. Members-only articles that Naver search links to are read through the search link's token; articles only members can find are skipped and reported in the run summary as skippedMembersOnly. A whole-cafe crawl can only return that cafe's public articles.
• View counts exist for cafe posts only. Naver does not publish a view count for blog posts found by search, so viewCount is empty on those rows. Whole-blog crawls fill it only for bloggers who make view counts public.
• Cafe posts carry no like count in the article data Naver serves; likeCount is filled for blog posts and for whole-cafe crawls.
• Some posts are genuinely short. Photo posts and one-line cafe posts have little text; that is the post, not a parsing gap.
• Dates are Korean time. A "day" window runs from 00:00 to 23:59 KST.

🧩 Related Actors

• Naver Map Scraper - Korean local businesses with website, Instagram and Naver blog links, phone and ratings.
• Lead List Cleaner - merge several runs into one deduplicated list.

🤖 Use with AI agents via MCP

The Actor is exposed over Apify's MCP server, so an agent can call it as a tool.

$claude mcp add apify --transport http https://mcp.apify.com --header "Authorization: Bearer YOUR_APIFY_TOKEN"

Then ask in words: "get last week's Naver Blog posts reviewing the Galaxy S26, with comments". Every option is a plain structured field, so the agent fills the input itself.

⚖️ Is it legal to scrape Naver Blog and Cafe?

The Actor reads only what Naver serves to any visitor without logging in: public blog posts, cafe articles that search exposes, and their public comments. It signs into nothing and bypasses no paywall. Posts and comments are written by individuals, so how you store and use them is your responsibility - Korea's PIPA and the GDPR both apply to personal data, and quoting or republishing content raises copyright questions. For research and analysis, aggregate rather than republish. Nothing here is legal advice.

❓ FAQ

Do I need a Naver API key?

No. There is no key, no developer account and no daily quota.

How do I get more than 1,000 posts for one keyword?

Set a date range. Each day in it is searched separately and each day can return up to 1,000 blog posts. Leaving the dates empty with Newest first also walks back day by day until your post limit.

Am I charged for posts I cannot read?

No. Members-only cafe articles that fail to open are skipped before anything is pushed, so they are never billed. The run summary counts them.

Are comments charged per comment?

No. It is one small charge per post whose comments were collected, however many comments it has.

Can it scrape a specific blogger or cafe?

Yes. Paste blog.naver.com/<id> or cafe.naver.com/<cafe> in Naver Blog / Cafe URLs and it returns their latest posts, up to your limit.

Should I search in Korean?

Yes. English keywords work but return a small fraction, because the posts are written in Korean.

🛠️ Support

Open an issue on the Actor's Issues tab with the run URL, the input you used and what you expected. That is the fastest route to a fix.

⭐ Rate this Actor

If this saved you time, please leave a rating on the Reviews tab. Review count is the main trust signal buyers use on the Store, and reviews are also how we decide what to build next. If something is broken or a field is missing, please open an issue first so it can be fixed - that is more useful to everyone than a low rating with no detail.