Japan Restaurants Scraper (Gurunavi)

Restaurant data from Rakuten Gurunavi (楽天ぐるなび, r.gnavi.co.jp) — one of Japan's largest restaurant guides — by prefecture, neighbourhood and cuisine, in a clean English schema: published business phone numbers, addresses with postal codes, budgets, opening hours, seat counts, smoking policy, private rooms, parking and accepted credit cards.

This Actor scrapes r.gnavi.co.jp listing pages (and optionally each restaurant's detail page) and returns structured JSON: restaurant name with katakana reading, cuisine genre, average lunch/dinner/party budgets as parsed yen values, nearest station and access line, street address, the shop's own published phone number, opening hours, closed days, total seats, facility facts and coordinates. Gurunavi is the restaurant guide where shops publish their real business phone numbers — which makes it the strongest Japanese source for restaurant B2B lead lists.

Unofficial tool. Not affiliated with or endorsed by Gurunavi, Inc. or Rakuten Group. This Actor collects publicly available factual business data only — it never collects review content, user identities or shop-authored promotional prose, and it never bypasses logins or CAPTCHAs. You are responsible for using the data in compliance with the source site's terms and the laws that apply to you (including GDPR/APPI where relevant); the published phone numbers are business contact information published by the restaurants themselves, not personal data — verify your own legal basis before using them for outreach.

---

Quick start — verified input

Copy-paste this into the Actor's input (JSON tab) or pass it via API — it was run successfully on 2026-09-02 (10 izakaya in Shibuya, Tokyo, listing data only, ~15 seconds):

{
  "prefecture": "Tokyo",
  "area": "shibuya",
  "genre": "izakaya",
  "maxItems": 10
}

Running with no input at all also works (defaults: Tokyo, all areas, all genres, 10 records). Add "detailFetch": true to fill address / phone / hours (one extra request per restaurant, ~1.2 s each).

Common input mistakes

Wrong	Right	Why
"prefecture": "Osaka-shi" / "Yokohama" / "Shibuya"	"prefecture": "Osaka" / "Kanagawa" / "Tokyo"	prefecture must be one of the 47 prefectures (English name). Cities and wards go in area, or nowhere.
"area": "Osaka" or "area": "Tokyo"	leave area empty (or "area": "areal2125")	area is a neighbourhood, not a city/prefecture. Take the code from a r.gnavi.co.jp listing URL.
"area": "Namba" (a non-Tokyo neighbourhood by name)	"area": "areal27xx" code copied from the site URL	Built-in English names exist for Tokyo only; every other neighbourhood works via its area code.
"genre": "Ramen restaurants"	"genre": "noodle" (or "ramen")	Use a Gurunavi genre slug (17 values — see the dropdown); ramen, yakiniku, italian, bar … are mapped automatically.
"prefecture": "東京都" or "13" or "tokyo" (via API)	"prefecture": "Tokyo"	The platform checks the 47 English names before the run starts; other spellings are rejected with the valid list.
"maxItems": "100"	"maxItems": 100	Numbers are numbers in JSON — the platform rejects a numeric string before the run starts.
"detailFetch": true with "maxItems": 100 and default maxApiRequests	add "maxApiRequests": 60 (or lower maxItems)	Each detail costs a request; when the budget runs out the run still succeeds but with complete: false.

Every invalid value stops the run immediately (before any request to r.gnavi.co.jp) with a message that lists the valid values.

Empty results?

A run that finishes with items: 0 and complete: true is not an error — the query simply matched nothing on Gurunavi. Typical causes: an area code from a different prefecture than prefecture (the area code wins — check the URL you copied), a very specific area + genre combination (e.g. a small areas… code plus buffet), or a genre that has few listings outside big cities. Widen the query (drop genre, or use the areal… large-area code) and re-run.

---

Who is this for?

• B2B suppliers and services — lead lists of restaurants by area and genre with the shop's own published phone number, address, seat count and budget positioning (POS systems, food suppliers, recruiting, insurance, delivery services).
• Market research and location intelligence — density, price points and facility profiles (private rooms, smoking policy, parking, seats) of restaurants by neighbourhood and cuisine.
• Travel, food-tech and recommendation products — English-schema Japanese restaurant data with reservation-grade facility details.

Looking for the Tabelog rating instead? Gurunavi publishes no public ratings — use the sibling Actor Japan Restaurants Scraper (Tabelog) for ratings, review counts and bookmark counts, and this Actor for phone-rich lead data; the two share an aligned schema (restaurantName / genre / budgetDinner / address / prefectureEn / phone / url) so the datasets join cleanly.

Sample output (dataset item)

{
  "restaurantName": "大衆居酒屋 酒場三九 渋谷店",
  "restaurantNameKana": "タイシュウイザカヤサカバサンキュウシブヤテン",
  "genre": ["居酒屋"],
  "budgetDinner": { "raw": "2,000円（通常平均）", "minJpy": 2000, "maxJpy": 2000 },
  "budgetLunch": { "raw": "1,000円（ランチ平均）", "minJpy": 1000, "maxJpy": 1000 },
  "budgetParty": { "raw": "3,500円（宴会平均）", "minJpy": 3500, "maxJpy": 3500 },
  "address": "東京都渋谷区宇田川町13-8ちとせ会館5F",
  "postalCode": "150-0042",
  "prefecture": "東京都",
  "prefectureEn": "Tokyo",
  "prefectureCode": "13",
  "nearestStation": "ＪＲ渋谷駅",
  "access": "ＪＲ渋谷駅 徒歩3分",
  "phone": "03-6427-8513",
  "businessHours": "月～木・日・祝前日・祝日 14:00～24:00 （L.O.23:30、ドリンクL.O.23:30） 金・土 14:00～翌4:00 （L.O.3:30、ドリンクL.O.3:30）",
  "closedDays": "年中無休",
  "totalSeats": 240,
  "smoking": "喫煙可（全席喫煙可）",
  "privateRooms": null,
  "parking": "無",
  "creditCards": ["VISA", "MasterCard", "JCB", "アメリカン・エキスプレス", "ダイナースクラブ"],
  "amenities": ["喫煙可", "お子様連れ", "食べ放題メニュー", "飲み放題メニュー"],
  "latitude": 35.6611418,
  "longitude": 139.6985508,
  "url": "https://r.gnavi.co.jp/9mpsba4d0000/",
  "gnaviShopNo": "9mpsba4d0000",
  "gnaviId": "gjf3506",
  "detailFetched": true,
  "source": "楽天ぐるなび (Gurunavi, r.gnavi.co.jp)",
  "sourceUrl": "https://r.gnavi.co.jp/",
  "license": "Publicly available data — unofficial tool; users are responsible for compliance with the source site's terms",
  "retrievedAt": "2026-08-25T09:12:44.000Z"
}

In cheap list-only mode (detailFetch: false, the default) each listing page request yields 30 records with name, genre, average budgets, access/nearest station and URL; address/phone/hours/facility fields are null.

The phone field is always the shop's own published business line (from the page's structured data) — never the Gurunavi 050 reservation tracking number shown in the on-page info table.

Input reference

{
  "prefecture": "Tokyo",
  "area": "shibuya",
  "genre": "izakaya",
  "maxItems": 60,
  "detailFetch": true,
  "maxApiRequests": 60
}

• prefecture (required): English name of one of the 47 prefectures exactly as in the dropdown ("Tokyo", "Osaka", "Kyoto" …). The Apify platform validates this list before the run starts, so "東京都", "13" or "tokyo-to" are rejected immediately with the list of valid values (the actor itself would accept them — this matters only for local apify run). Cities are not prefectures — "Yokohama" is rejected; use "Kanagawa".
• area (optional): a Gurunavi area code copied from any r.gnavi.co.jp URL — areal2125 (large area), aream… / areas… (medium/small) — this works for every neighbourhood in Japan. Convenience English names are built in for famous Tokyo areas (shinjuku, shibuya, ginza, ikebukuro, roppongi, ueno, asakusa, akihabara, ebisu, shinagawa, kichijoji …; Shibuya-ku and 渋谷 also work). Unknown free text is refused with the list of valid values instead of guessing the wrong neighbourhood.
• genre (optional): one of the 17 Gurunavi genre slugs (izakaya, japanese, sushi, roastmeat, western, buffet, curry, chinese, continental, hotpots, ethnic, noodle, beerandcuisine, cafe, entertainment, casual, others) with common aliases (ramen→noodle, yakiniku→roastmeat, italian/french→western, bar→beerandcuisine) and Japanese labels (居酒屋, 寿司 …). Anything else is refused before the first request, listing all valid genres.
• maxItems (default 10, max 1000) / maxApiRequests (default 15, max 60) / detailFetch (default false) / proxyConfiguration (default none).

Pricing

Pay per result — see the pricing tab. A listing-only run is cheap (30 records per upstream request); detailFetch: true costs one extra upstream request per restaurant and fills the business-info fields.

FAQ

Where does the data come from? Live from public r.gnavi.co.jp listing and shop pages at run time. Every record embeds the source, source URL, a license note and the retrieval timestamp.

Is this an official Gurunavi/Rakuten service? No. Unofficial, independently built and maintained, not affiliated with or endorsed by Gurunavi, Inc. or Rakuten Group. Gurunavi is a trademark of its owner. (The official Gurunavi web API stopped accepting new registrations in 2021 — this Actor reads the public website instead.)

Does it collect reviews or user data? No, by design. The parser deletes any review array from the page's structured data before reading anything, never reads user content, and also skips shop-authored promotional prose (descriptions, photo captions) and paid ad placements — automated tests assert that none of it can ever appear in the output. Only factual business data survives. Gurunavi publishes no rating scores, so there are no rating fields — see the Tabelog sibling Actor for ratings.

Are the ratings missing? Gurunavi does not publish public rating scores. If you need ratings, run the sibling Japan Restaurants Scraper (Tabelog) — its schema is aligned with this one so you can join on name/address/coordinates.

Why did my run stop before maxItems? The Actor enforces a hard per-run request budget (maxApiRequests) with strictly serial requests ≥1.2 s apart — polite by construction. The run summary (RUN_SUMMARY key-value record) reports complete: false with the reason whenever a budget truncated the result set. Raise maxApiRequests or narrow the query.

What if r.gnavi.co.jp blocks the datacenter IP? The Actor performs no block evasion (no header spoofing, no CAPTCHA solving, no rotation tricks). If the runtime IP is refused it fails visibly with the HTTP status. Setting proxyConfiguration (e.g. Apify automatic proxy) usually resolves it.

Is the whole of Japan covered? Yes — any prefecture, and any neighbourhood via its Gurunavi area code. What one run returns is your query's result set subject to maxItems and the request budget, both reported honestly in the run summary.

Search terms this Actor answers

gurunavi scraper · gnavi scraper · Gurunavi API · Japan restaurant database · Japan restaurant leads · restaurant phone numbers Japan · Tokyo dining data · japan restaurant scraper · izakaya list Tokyo · Japanese restaurant contact list · ぐるなび データ · restaurant B2B leads Japan · Japan food service market data

More Japan data Actors by the same developer

• Japan Restaurants Scraper (Tabelog) — the same aligned schema with the famous Tabelog rating, review counts and bookmark counts.
• Japan Government Tenders & Awards — tender notices and award results from the kkj.go.jp portal.
• Japan Medical Facilities — every registered hospital, clinic, dental clinic, and pharmacy (MHLW open data).
• Japan Care Facilities — long-term care providers across 35 service types (MHLW open data, CC BY).
• Japan Government Subsidies & Grants — open subsidy calls from the official jGrants API.
• Japan Construction & Real-Estate Licenses — licensed contractors and brokers from MLIT's official register.
• Japan Financial Licenses & Registrations — FSA-licensed banks, securities firms, crypto exchanges and more.

---

Data source: r.gnavi.co.jp public pages. Unofficial; not affiliated with Gurunavi, Inc. or Rakuten Group. Factual business data only — no review content, no user data, no promotional prose. Users are responsible for their own compliance with the source site's terms and applicable law.