Rakuma Japan Sold and Asking Prices — Median, Range

Unofficial — independent tool, not affiliated with, endorsed by, or sponsored by Rakuma. It reads only publicly visible pages. Support, reliability guarantees and the full disclaimer are at the bottom of this page.

What it does: Look up any word on Rakuma, Rakuten's flea market in Japan, and get both what it sold for and what it is listed for today.

You enter: one or more search words, in Japanese (ポケモンカード リザードン) or English (leica m6).

You get: from the listings read (40 a page, 3 pages by default, spread over the whole result list): typical price and lowest–highest range in yen and US$, the sold and the still-listed price apart, how many already sold, the shop-versus-private mix, Rakuten point-back.

Price: $0.02 per keyword. +$0.002 per row if you also want the list. No results = no charge.

Example: enter leica m6 → 398 matches on Rakuma, 118 listings read from pages 1, 6 and 10, of which 35 are really that camera · typical ¥355,555 · range ¥6,780–3,241,000 · the 15 already sold typically ¥290,000, the 20 still listed ¥576,000 · 43% of them listed by shops (real run, 2026-09-08)

Unofficial — not affiliated with Rakuma. Reads public pages only.

Pricing — $0.02 per keyword

Event	Price	When
Keyword price summary	$0.02	Per keyword measured
Individual listing	$0.002	Only if you enable Include individual listings

A default run (1 keyword, summary only) costs $0.02. You pay per keyword; there is no monthly fee. A keyword that returns zero listings of your item is never charged — and that matters more here than on most sites, because Rakuma answers a word it does not know with unrelated goods rather than with an empty page (see The word that finds nothing below). When a keyword is not charged, its listing rows are not produced or charged either. About 12 keywords fit one run at the default depth — after that the run's sampling budget is spent and the remaining words are measured on their first page only (truncatedForTimeLimit: true) — and a run also stops at the $3.00 Maximum cost per run. Raise either in the run options, or split the list across runs.

Input

Field	Example	Notes
keywords	["leica m6"]	One or more search words. Japanese and English both work well (ポケモンカード リザードン, leica m6). Each word costs $0.02
maxItemsPerKeyword	120	40–800. Rakuma serves 40 listings a page and has no page-size setting, so this ÷ 40 is the number of pages read per word — 120 is three. Also caps the listing rows
titleMatch	"all"	all (default) keeps listings whose title carries every word you typed, any keeps those carrying at least one, off keeps everything returned. all widens to any by itself when it keeps none; the row says which was used in matchMode
includeSoldOut	true	On by default: already-sold items count towards the headline price. Off prices only what is still on sale — sold items are still counted and still get soldPriceJpy
includeIndividualItems	false	Enable to also get one row per listing read (+$0.002 each)
convertToUsd	true	Adds US$ figures at the current exchange rate. A failed lookup never fails the run

Sold price and asking price, from the same search

Most Japan flea-market tools give you one or the other. Mercari's public search shows sold items; PayPay Flea's shows only what is on sale. Rakuma's default search returns both, mixed together, and marks the sold ones — so one read gives you:

• soldPriceJpy — what buyers actually paid, over the sold listings among your matches
• askingPriceJpy — what sellers are asking right now, over the rest
• priceJpy — both together (or only the still-listed ones if you turn includeSoldOut off; soldOutIncludedInPrice says which)

Two warnings about reading them:

• soldShare is not a speed of sale. Nothing on a Rakuma card says when it sold, so this is the share of the listings you read that had already gone — not how fast the item moves. It also moves with how deep into the result list you look.
• The sold and asking figures can cross over. In the measured run below the sold price (¥290,000) is below the asking price (¥576,000), because the recommended order puts expensive unsold stock near the top. That is a fact about the listings, not a discount rate — do not divide one by the other.

The word that finds nothing

Rakuma never returns an empty page. Type a word it does not know and it answers with a few hundred unrelated goods and a perfectly healthy count — measured 2026-09-08, zzzqqxnotarealproduct returned 199 matches and forty pairs of earrings. Reported naively that is a confident price for the wrong thing, charged for.

So every listing read is checked against the words you typed, and the row tells you what happened:

keywordStatus	What it means
ok	Listings matched your words and carried prices — this row has figures, and it is charged
not_found	Rakuma returned listings, but none of them is your item. hint names the next thing to try. Not charged
unknown	Listings matched your words but none showed a price at all. hint says so. Not charged

One case sits outside the table: with Include sold items in the price turned off, a word whose every match has already sold keeps keywordStatus: "ok" — the prices are real and are right there in soldPriceJpy — but the headline priceJpy comes back empty, hint says to turn the switch on, and the row is not charged.

matchedListings is how many of sampledListings counted as your item, so you can always see how much of what came back was noise.

Output example (type: "price_summary")

Measured on 2026-09-08 (real run, {"keywords": ["leica m6"], "maxItemsPerKeyword": 120, "titleMatch": "all", "includeSoldOut": true, "includeIndividualItems": false, "convertToUsd": true}). The whole record, nothing shortened. Three pages were read — 1, 6 and 10 of the ten pages the result set has — and all 35 matching listings turned out to sit on page 1: Rakuma returns 398 listings for those words, but only 35 of them are the camera.

{
  "type": "price_summary",
  "keyword": "leica m6",
  "keywordStatus": "ok",
  "matchMode": "all",
  "currency": "JPY",
  "priceIncludesTax": true,
  "totalListingsFound": 398,
  "totalListingsShownOnPage": 300,
  "reachableListings": 398,
  "pagesSampled": [1, 6, 10],
  "sampledListings": 118,
  "matchedListings": 35,
  "pricedListings": 35,
  "soldOutIncludedInPrice": true,
  "priceBasis": "sample",
  "priceJpy": { "min": 6780, "p25": 18400, "median": 355555, "p75": 611000, "max": 3241000, "average": 507008 },
  "priceUsd": { "min": 43.91, "p25": 119.16, "median": 2302.57, "p75": 3956.84, "max": 20988.72, "average": 3283.38 },
  "exchangeRateJpyUsd": 0.006476,
  "soldPriceJpy": { "min": 6780, "p25": 93950, "median": 290000, "p75": 469000, "max": 1299000, "average": 329186 },
  "askingPriceJpy": { "min": 6780, "p25": 21200, "median": 576000, "p75": 829500, "max": 3241000, "average": 640374 },
  "soldListings": 15,
  "soldShare": 0.4286,
  "shopSellerListings": 15,
  "shopSellerShare": 0.4286,
  "pointBackListings": 10,
  "pointBackMaxRatePct": 3,
  "priceCrossCheckOk": true,
  "relatedKeywords": ["leica フィルムカメラ 一眼レフ", "leica m6 bundled_item", "s2", "s2 新品", "rolex 1908", "rolex", "dior homme"],
  "hint": null,
  "sourceUrl": "https://fril.jp/s?query=leica+m6",
  "checkedAt": "2026-09-08T17:29:20.373799+00:00",
  "truncatedForTimeLimit": false
}

With Include individual listings on, each listing read is also returned as its own row: itemId, url, name, priceJpy, isSoldOut, sellerType (shop or person), pointBackRatePct, categoryId and matchesQuery.

What this Actor does not do

• No numbers covering every listing. Rakuma's search cannot be ordered by price — that setting is closed to automated readers by the site's own robots.txt — so no position in the result list stands for the whole set. Every price figure covers the listings actually read and says so (priceBasis: "sample")
• No filtering by price, category, brand, condition or "sold only". All of those are the same closed settings. What you can narrow is the search word itself and titleMatch
• No sale dates, so no time-to-sell. A Rakuma card carries no date; soldShare is a share of what was read, not a speed
• No seller identity. The markup carries the seller's user number; this Actor throws it away and keeps only whether the seller is a business or a private person. No names, no contact details, no seller pages
• No images and no description text. Only the listing's own facts (name, price, sold mark, seller type, point-back, category) and the aggregate figures
• No login-only data. Everything comes from public search pages, with no account and no cookie
• Nothing is stored. Every run reads the site live; nothing is kept between runs

Notes on the data

• The count is exact. totalListingsFound comes from Rakuten's own counter, not from the rounded 「約300件中」 the page prints — that rounded number is reported separately as totalListingsShownOnPage purely so a divergence between the two shows up instead of passing silently. Verified 2026-09-08: a search declaring 3,158 matches ended on page 79 with 38 cards, and 78 × 40 + 38 = 3,158 exactly
• Counts drift within the hour (763,874 → 763,974 on the same word an hour apart), which is why every row is stamped with checkedAt
• Only the first 4,000 matches are readable. Rakuma stops at page 100; reachableListings is totalListingsFound capped at 4,000
• Pages are spread, not taken from the top. The pages read are stretched evenly over the whole result list, because both the price and the sold share change sharply with depth: measured 2026-09-08 on a 3,158-match search, page 1 was 1 sold out of 40 at a typical ¥97,780 and page 79 was 32 of 38 at ¥8,900. pagesSampled lists exactly which pages the row is built from
• Titles are cut short by the site, so a word sitting late in a long listing name can be missing from the card. That is why titleMatch: "all" widens to any on its own when it would otherwise keep nothing, and why the row reports the matchMode it ended up using
• "Sold" means the listing carries Rakuma's own SOLD OUT ribbon. Flea-market prices are fixed, so that price is what the item actually went for (or its last price, if the seller marked it down first)
• Prices are in yen and include tax, and each one is read twice — from the analytics attribute and from the price the page prints — so that a change to Rakuten's tags is reported (priceCrossCheckOk: false) rather than quietly returning a wrong number
• pointBackMaxRatePct is the highest Rakuten point return, in percent, among the matching listings. It is a discount in points, not off the price
• relatedKeywords is Rakuma's own 「他のお客様はこちらも検索」 list for your word — useful for finding the spelling Japanese sellers actually use. It appears on the first page only, so it can be empty
• Pages are UTF-8 and fully server-rendered; requests are spaced at least 1.5 seconds apart and a run carries a wall-clock budget. If the budget is reached, later keywords are measured on fewer pages and their rows carry truncatedForTimeLimit: true

If something goes wrong

• Wrong number or a failed run? Open a ticket on the Issues tab. I read every one and reply within 2 business days (Japan time).
• You never get a fake "empty" result. If the site can't be read, the run fails and says so.
• No results = no charge. You only pay for results you actually get.
• Checked every week. An automatic test runs this tool weekly; if the site changes, I fix it.
• Public pages only. No login, no personal data, and it goes easy on the site.

More tools by the same author

• Mercari Japan Sold Prices — What Items Really Sell For
• Yahoo! Auctions Japan Sold Prices — Median, Range, Bids
• Japan Used Camera & Lens — Dealer Price vs Sold Price
• Japan Figure & Gunpla — New Price vs Used Sold Price
• Used iPhone & Android Prices Japan — Dealer vs Sold
• Yahoo, Mercari & PayPay Japan — Same Item, 3 Prices
• Mercari & Yahoo Japan Demand — Sold vs Still Listed
• Pokemon Card Japan Sold Prices — Raw vs Graded per Card

All tools (Japan marketplaces, real estate, jobs, racing, prediction markets): https://apify.com/jpmarketdata

Disclaimer

Unofficial, independent tool — not affiliated with, endorsed by, or sponsored by Rakuma. Product names and logos belong to their owners and only say where the data comes from. Data is read from public pages, for market research; check before you act on it.