Start free

Platforms MCP server

MCP server · 2 tools · 4 endpoints

Naver MCP server

Blog, cafe, local and shopping search across the Korean web, for agents that need more than Google can see there.

Get a free key Connect in a minute

100 credits on signup · No card · Sign in with OAuth or paste a key

scrapersocial-naver mcp.scrapersocial.com/naverconnected
you

Search Naver blogs for <Korean keyword> and summarise what people are writing about it.

agent
calls list_endpoints free
server
endpoints [{ path, input, credits }] · /v1/naver/blog-search is 2 credits
agent
calls call_endpoint 2 credits · the endpoint's own cost
{ "path": "/v1/naver/blog-search", "query": "<keywords>" }
server
returns
post object · the full post record
Shape only. Keys follow the tool's summary; a real call returns the full record. 2 calls · 2 credits

Connect

One URL. Your client already knows what to do with it.

The first call opens a sign-in. New accounts get 100 credits, no card. Every client below shares one account and one balance.

Get a free key
server url https://mcp.scrapersocial.com/naver
claude.aiOAuth

Settings → Connectors → Add custom connector, then paste the server URL.

mcp.scrapersocial.com/naver

Web, desktop and mobile. Nothing else to paste.

Claude Codeone command

Run this in a terminal.

$ claude mcp add --transport http scrapersocial-naver https://mcp.scrapersocial.com/naver

Approve the sign-in prompt on first use.

Cursorone click

Installs the server and opens the MCP panel.

Add to Cursor
or paste the JSON into .cursor/mcp.json
{ "mcpServers": { "scrapersocial-naver": { "url": "https://mcp.scrapersocial.com/naver", "headers": { "Authorization": "Bearer sk_live_..." } } } }
Other clients · VS Code, Windsurf, Gemini CLI, anything that speaks MCP
VS Codeone click or JSON
Add to VS Codeor run MCP: Add Server, choose HTTP and use:
{ "name": "scrapersocial-naver", "type": "http", "url": "https://mcp.scrapersocial.com/naver" }
Windsurf~/.codeium/windsurf/mcp_config.json
{ "mcpServers": { "scrapersocial-naver": { "serverUrl": "https://mcp.scrapersocial.com/naver", "headers": { "Authorization": "Bearer sk_live_..." } } } }
Gemini CLI~/.gemini/settings.json
{ "mcpServers": { "scrapersocial-naver": { "httpUrl": "https://mcp.scrapersocial.com/naver", "headers": { "Authorization": "Bearer sk_live_..." } } } }
Any clientStreamable HTTP · OAuth 2.1 or Bearer
$ npx add-mcp https://mcp.scrapersocial.com/naver

One account and one balance across every client and the every-platform hub at mcp.scrapersocial.com.

What it does

What your agent can do with Naver

If the market you are researching is Korea, Google data will not represent it. Naver is where the audience and the content are, and an agent that cannot read it is guessing. Naver’s spoke, https://mcp.scrapersocial.com/naver, gives an MCP client four Naver searches that currently serve, each at 2 credits per result.

No capability tools exist on this spoke; the generic pair does all the work. A free list_endpoints call names the paths that can answer right now. call_endpoint runs them with a query: /v1/naver/blog-search for blog posts, /v1/naver/cafearticle-search for posts inside Naver Cafe communities, /v1/naver/local-search for business listings, and /v1/naver/shop-search for product listings and prices. Each is a list, so limit and cursor work the way they do everywhere else on ScraperSocial.

Query in Korean. The index is Korean-language and Korean-market, and a Korean query returns substantially better results than an English one; have your agent translate the term first if the user wrote it in English. Naver is a portal rather than a social network, but the cafe communities are among the most candid places a brand gets discussed in Korea, which is why this spoke sits next to Reddit and X on the hub.

Four jobs it finishes on its own

01

Brand monitoring in Korean communities

Run /v1/naver/cafearticle-search for a product name and have the model summarise what members are saying. Cafe posts are often franker than reviews, and twenty of them cost 40 credits.

02

Korean price checks

/v1/naver/shop-search returns product listings with prices for a query, so an agent can report what a product sells for on the Korean market without anyone opening a browser in Korean.

03

Local listings for a Korean city

Ask for a category in a district and /v1/naver/local-search returns the businesses. Pair it with the Google Maps spoke to see which listings exist in both places and which only Naver knows about.

04

Blog coverage research

Naver blogs carry much of Korea's long-form consumer writing. /v1/naver/blog-search finds posts on a topic so the model can read how a product or trend is being written about.

Limits · read before you connect

What this does not do

Written by a person, not generated. If something here is wrong, tell us.

Four paths serve today. News, book, encyclopedia, Q&A, document, image and general web search are not offered on this spoke yet, and neither is a combined brief; list_endpoints shows only the four that answer. Every working path is a search; there is no profile, post or comment endpoint. Each path pages its results: 20 per request unless told otherwise, a maximum of 50, and next_cursor in meta for the remainder. Results are Korean-language. No Naver account is involved and nothing writes. An empty search is still 1 credit.

Tools

2 tools, and 4 endpoints behind them

Naver has 4 endpoints, so there is nothing to wrap. list_endpoints tells the agent the path; call_endpoint runs it at the endpoint's own price.

Full credit table →

List every Naver endpoint

list_endpoints

Browse every Naver endpoint, its input and its credit cost. Free, and calls nothing upstream.

free

Call any Naver endpoint by path

call_endpoint

Run any Naver endpoint by its path, so nothing Naver offers is out of reach of an agent.

endpoint's own cost
What list_endpoints returns
/v1/naver/blog-search · input: query 2 credits / post /v1/naver/cafearticle-search · input: query 2 credits / post /v1/naver/local-search · input: query 2 credits / place /v1/naver/shop-search · input: query 2 credits / product

Pricing

100 free credits is 50 Naver posts.

Then $5 a month for 1,000 credits, or $4 per 1,000 as a top-up. No card to start. Every call costs 2 credits.

50 Naver posts/v1/naver/blog-search · 2 each
Get a free key 100 credits on signup · $5 for 1,000 · full credit table

Prompts

Things to ask once it is connected

Copy one, swap the placeholder, paste it into your client.

Search Naver blogs for <Korean keyword> and summarise what people are writing about it.

What are Naver Cafe members saying about <product name in Korean>? Pull the latest posts and summarise the mood.

Find <category> businesses in <Seoul district> on Naver and list them with addresses.

What does <product> cost on Naver Shopping? Show the price range across sellers.

Translate <English term> into Korean, search Naver blogs for it, and tell me the top themes.

Questions

Before you connect

What does a Naver search cost? #

Each result is 2 credits on all four working paths, twenty results a page and no more than 50 in one call. The free 100 credits cover fifty results; a $5 month covers five hundred.

Do I need to search in Korean? #

You should. The content is Korean, and a Korean query matches far more of it. A good pattern is to have the agent translate the user's term before calling the endpoint.

Why doesn't news or image search work? #

They're not available yet. list_endpoints hides anything that can't serve, and call_endpoint declines it without charging. Blog, cafe, local and shopping are what answer today.

Is there a tool like search on this spoke? #

No capability tools at all. Everything goes through call_endpoint by path, which is how the agent gets at the four searches once list_endpoints has shown it the paths.

Is ScraperSocial connected to Naver? #

No. Naver has no part in it; ScraperSocial is an independent service not affiliated with Naver Corporation, and it reads publicly available Korean web data only.