Naver Places Reviews Scraper API
Naver Places review pages carry the signals local-business teams care about: what customers wrote, who wrote it, when, and how the listing's aggregate rating and review count look.
This page explains how a managed API call turns a public Naver Places URL into structured review data, which request options change the output and cost, and where the boundary sits between public-page extraction and content behind a login.
What a Naver Places review call returns
Each call targets a public Naver Places review page and returns the same core field set. The fields below are the ones the API is documented to produce; treat them as the schema you design around, not as a guarantee that every listing populates every field.
- Business name: the listing the reviews belong to.
- Aggregate star rating: the listing-level rating shown on the page.
- Review count: how many reviews the listing reports.
- Individual review text: the body of each review.
- Reviewer handle: the display name attached to the review.
- Review timestamp: when the review was posted.
If you need a field outside this set, use extract_rules with CSS selectors or ai_extract_rules for AI-driven extraction. Anything that renders on the public page is a candidate; anything behind a login is not.
How one call is put together
The request path is short by design. You create an account to get an API key and starting credits, then send the Naver Places URL with that key. The service renders the page in a real browser, rotates the IP, and returns the response your scraper acts on. You do not run headless Chrome or manage a proxy pool yourself.
Optional parameters extend the call when the page needs more than a plain fetch:
- render_js for JavaScript-rendered content.
- stealth_proxy when CAPTCHA or anti-bot handling is required.
- country_code for geo-targeted requests.
- ai_extract_rules for natural-language extraction into structured JSON.
SDKs are published for Python, Node.js, Go, PHP, Ruby and Java, and plain cURL works too. Pick the stack your codebase already uses; the call shape is the same.
Field map: default output vs extraction rules
Use this map to decide whether the default response is enough or whether you need to declare selectors or an AI extraction rule.
- Business name — default response. Use extract_rules only if you need a normalized variant.
- Aggregate star rating — default response. Numeric; watch for locale formatting on the page.
- Review count — default response. Useful as a denominator for sentiment or coverage calculations.
- Review text — default response. This is the field most pipelines feed into sentiment or topic models.
- Reviewer handle — default response. Treat as public display data; do not attempt to enrich it into personal identity.
- Review timestamp — default response. Normalize to a single timezone before joining across listings.
- Anything else on the page — extract_rules (CSS selectors) or ai_extract_rules (natural-language rule).
A practical pattern is to keep the six default fields as your stable schema and add extract_rules fields only when a downstream report needs them. That keeps the pipeline resilient when Naver changes markup outside the core review block.
Worked example: a review-monitoring loop
Suppose you monitor a set of Korean local-business listings. On a schedule, you send each listing's public Naver Places review URL with your API key and render_js enabled. For each response you keep the six default fields and write them into a table keyed by business name plus review timestamp.
On the next run, you compare review counts and timestamps against the previous snapshot. New rows are the reviews posted since the last run; changed aggregate ratings are the listing-level movement. From there you can compute a simple sentiment score over the new review text and chart it against the aggregate rating to see whether the listing-level number is tracking the recent review tone.
Two failure distinctions matter here. A request that returns a page but no review block is usually a rendering or selector issue, not a legal one; check render_js and your extract_rules. A request that targets post-login content is out of scope for this pattern regardless of options. Keep the two separate in your runbook so on-call engineers do not chase the wrong cause.
Credit cost by option combination
Credit consumption depends on which options a request enables. The figures below are the documented per-request costs; they are the numbers to use when sizing a monthly budget.
- Base request: 1 credit.
- JS rendering enabled: 5 credits per successful request.
- Premium proxy without JS rendering: 10 credits per successful request.
- Premium proxy with JS rendering: 25 credits per successful request.
- stealth_proxy: 75 credits per successful request.
API credits are valid for one month and do not roll over. Concurrent requests are the number of requests that can be sent at the same time. Plan tiers set both the credit allowance and the concurrency ceiling; the plan table on this page lists the current tiers and their limits.
Decision checklist before you build
Run through these before committing to a pipeline shape.
- Is the target content public? If it sits behind a login, this pattern does not apply; stop and reassess.
- Do you need the six default fields only, or extra fields? Extra fields mean extract_rules or ai_extract_rules, which changes your test plan.
- Will the page need JavaScript rendering? If yes, budget 5 credits per successful request rather than 1.
- Will the page need a premium proxy? Without JS rendering that is 10 credits per successful request; with JS rendering it is 25.
- Will the page trigger CAPTCHA or anti-bot handling? If yes, stealth_proxy at 75 credits per request is the relevant cost.
- Do you need geo-targeted requests? If yes, country_code is the parameter to set and the plan must cover the concurrency you need.
- Have you reviewed the target site's terms of service? Public-facing pages can usually be scraped, but terms review is part of the build, not an afterthought.
- Do you have a snapshot-comparison step? Without one, you will re-ingest the same reviews every run.
Plan reference
The plan table on this page lists the current tiers, monthly prices, credit allowances and concurrency limits. Prices are exclusive of VAT. A free starting allowance is available without a card so you can test the call shape before choosing a tier. Use the table to match a tier to your monthly request volume after applying the per-request credit costs above.
| Plan | Price / month | API credits | Concurrent requests | Priority email support | Dedicated account manager | Team management |
|---|---|---|---|---|---|---|
| Hobby | $19 | 75,000 | 25 | Not included | Not included | Not included |
| Freelance | $49 | 250,000 | 50 | Not included | Not included | Not included |
| Startup (Recommended) | $99 | 1,000,000 | 100 | Not included | Not included | Not included |
| Business | $249 | 3,000,000 | 200 | Included | Included | Included |
| Business + | $599 | 8,000,000 | 400 | Included | Included | Included |
| Enterprise 14 | $999 | 14,000,000 | 500 | Included | Included | Included |
| Enterprise 24 | $1,599 | 24,000,000 | 600 | Included | Included | Included |
| Enterprise 41 (Recommended) | $2,399 | 41,000,000 | 700 | Included | Included | Included |
| Enterprise 69 | $3,599 | 69,000,000 | 800 | Included | Included | Included |
| Enterprise 120 | $5,799 | 120,000,000 | 900 | Included | Included | Included |
Prices are exclusive of VAT. Credit allowances are monthly and do not roll over. Requests start at 1 credit each. Concurrency is the number of simultaneous requests.
Frequently asked questions
Do I need a Naver account to scrape Naver Places reviews?
No external account is required. You sign up for an API key, which comes with 1,000 free credits and no card, then send the target URL. The service handles the browser and proxy side.
Is scraping Naver Places reviews legal?
Public-facing pages can usually be scraped, but review the target site's terms of service before deploying and avoid post-login content.
What data can I extract from Naver Places reviews?
Anything that renders on the public page. The default response covers business name, aggregate rating, review count, review text, reviewer handle and review timestamp. For other fields, use extract_rules with CSS selectors or ai_extract_rules for AI-driven extraction to get structured JSON back.
How do I start scraping Naver Places reviews?
Sign up, grab your API key, and send a GET request with the target URL. The service renders the page, rotates proxies and returns the data. You get 1,000 free credits to test with.
Which programming languages can I use?
Any. SDKs are published for Python, Node.js, Go, PHP, Ruby and Java, and plain cURL works too. Pick whatever your stack already uses.
How much does it cost to scrape Naver Places reviews?
Signup includes 1,000 free credits with no card required. Paid plans start at $19 per month for higher volume and concurrency. Per-request credit cost depends on options: 5 credits for JS rendering, 10 for premium_proxy without JS rendering, 25 for premium_proxy with JS rendering, and 75 for stealth_proxy.