Mercari Japan Price Summary — Sold Out & Live

What it does: Checks Mercari Japan sold-out or on-sale listing prices and returns a price summary, with optional supporting item rows. Sold-out labels do not verify payment.

You enter: Keywords, e.g. WH-1000XM5; status; sample limit; optional JPY price band and evidence level.

You get: Median, middle 50% range, lowest–highest, average, sample count and approximate days to sell. Choose no item rows, up to 10 representative rows, or all sampled priced rows.

Price: $0.02 per nonempty keyword/status summary. Representative or full evidence adds $0.002 per returned item. both can cost $0.04 in summaries per keyword. No priced results = no charge.

Example: For a sold-out price benchmark, use keywords: ["WH-1000XM5"], itemStatus: "sold_out", evidenceMode: "summary" and a JPY price band. A nonempty summary costs $0.02; inspect the returned sample count and price range before deciding.

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

Live API beta: Check a single keyword and receive summary JSON directly. Open the Server tab or follow the HTTP quick start.

Choose the product for your job

sold_out is the default listing status. Choose on_sale for current asking prices, or both for separate live and sold-out summaries ($0.04 per keyword). The source labels do not independently prove payment.

Use this Actor for calculated price statistics, from 30–600 sampled rows per keyword/status, with optional JPY price bounds. summary is the default evidence mode; representative adds up to 10 rows across the sampled price range, and full returns every sampled priced row. Summary costs $0.02 per keyword/status; evidence adds $0.002 per returned row.

If your job is to ingest and process the individual items yourself, choose Mercari Japan Listings Scraper — Bulk Item Data: up to 2,000 rows per keyword, live or sold-out status selection and JPY price bounds, at $0.0015 per delivered row with no keyword fee. The source status is retained on each evidence row. Sold-out search can include items still trading; the published price is not independent proof of payment or a completed transaction.

Repeat a saved price watchlist

Save fixed keywords, itemStatus, maxItemsPerKeyword, priceMinJpy / priceMaxJpy and evidenceMode as an Apify Task; set convertToUsd: false for a JPY series. Schedule the Task at the cadence you need, retain each run's dataset and compare price_summary rows by keyword, status and checkedAt, alongside sampledListings and priceJpy.

Keep the same settings between runs. A changing sample can move the median without a same-item price change. sold_out is a source label and does not independently prove payment or a completed transaction. Each scheduled run uses the same summary and optional-evidence charges.

日本語ガイド — メルカリの売り切れ相場を調べる

概要: メルカリの売り切れ(SOLD)商品をキーワード検索し、売り切れとして公開されている掲載価格 — 中央値・平均・真ん中50%の価格帯・最安〜最高 — と「売れるまでの日数」の中央値(概算)を返します。1キーワード・状態につき1行です。売り切れ表示は入金や取引完了を独立確認したものではありません。出品前の相場チェック、仕入れ・せどりの判断、eBay など海外市場との価格差の確認に使えます。

使い方: 入力欄にキーワードを入れて Start を押すだけ。日本語キーワードが最も精度が高く、複数キーワードの一括チェックもできます。

{
    "keywords": ["ポケモンカード リザードン psa10"],
    "itemStatus": "sold_out",
    "evidenceMode": "representative",
    "maxItemsPerKeyword": 120,
    "convertToUsd": true
}

料金: 価格サマリーは1キーワード・1状態あたり $0.02(約3円)。根拠個票は、価格帯全体をカバーする代表10件または取得した全件から選べ、返した個票1件につき +$0.002 です。both は売り切れ・出品中を別々に計算します。価格のある結果が0件なら課金されません。

How Mercari sold-out price research helps

Compare source-labelled sold-out listing prices with current asking prices when researching trading cards, anime figures, film cameras, watches, fashion or retro games. Search by a specific model and price band to reduce accessories, bundles and condition differences. This Actor returns a separate summary row for each keyword/status:

• Price a collection before you buy or sell — the median and the middle 50% range of source-labelled sold-out listings
• Spot arbitrage — compare sampled Japan listing prices against another market, allowing for condition, fees, shipping and exchange rates
• Track liquidity — medianDaysToSell estimates time between listing creation and the last source update; it is not a verified sale interval
• Feed your AI agent or spreadsheet — clean JSON, USD conversion included

Price — know your cost before you run

You pay per keyword; there is no monthly fee.

Evidence level	What is returned	Price
Summary	The price answer only	$0.02 per keyword/status
Representative	Summary plus up to 10 price-bearing listings spread across the sampled price range	+$0.002 per returned listing
Full	Summary plus every price-bearing listing sampled	+$0.002 per returned listing

• Summary is the default and keeps the original flat $0.02 price per keyword/status.
• Representative is recommended when you need supporting listing examples. Its rows are selected across the full sampled price distribution, not taken from the first 10 results.
• Full is for audit or modelling work that needs every sampled listing carrying a price.
• A search with no price-bearing result is never charged.

HTTP price lookup beta

Check one item type while deciding what to list or bid: send one keyword to POST /price-check and receive the price summary directly as JSON. This Actor's live beta is enabled. You need an Apify account and your own API token from Settings → Integrations. Set APIFY_TOKEN in your environment, then run:

curl --fail-with-body --request POST \
  "https://jpmarketdata--mercari-japan-price-checker.apify.actor/price-check" \
  --header "Authorization: Bearer $APIFY_TOKEN" \
  --header "Content-Type: application/json" \
  --data '{"keyword":"WH-1000XM5","itemStatus":"sold_out","priceMinJpy":12000,"priceMaxJpy":40000}'

Measured on 2026-10-13 (owner-account test, response excerpt):

{
  "keyword": "WH-1000XM5",
  "status": "sold_out",
  "keywordStatus": "ok",
  "sampledListings": 120,
  "priceJpy": {"min": 16000, "p25": 19949, "median": 21000, "p75": 23125, "max": 36000, "average": 21884},
  "mixedKeywordLikely": false
}

Two sequential warm owner-account lookups took 1.25 and 1.45 seconds. These are observations, not a latency guarantee; startup and source availability can add time. Use OpenAPI JSON with the same authorization header to inspect the complete API specification. A request to GET / checks readiness without a Mercari search or event charge.

The HTTP interface accepts one keyword (1–120 characters), itemStatus (sold_out or on_sale, default sold_out), maxItemsPerKeyword (30–120, default 120), optional priceMinJpy (zero or greater) / priceMaxJpy (1 or greater), and convertToUsd (default false). It returns the same price_summary fields: median, middle 50% range, lowest–highest, sample size, keyword-mixing flag, approximate days to sell and check time. It samples one search page; it is not a complete sales history or an exact appraisal of a particular edition or condition. on_sale returns asking prices. Days to sell are estimated from source timestamps, not a verified sale date.

HTTP returns summary only, at the existing $0.02 per nonempty keyword/status result. A normal zero match returns the search hint without charge; acquisition and billing failures return an HTTP error instead of an empty success. Results are delivered directly to the caller and are not saved in shared result storage. Use normal Input runs for multiple keywords, both, representative listings or full evidence. Authentication uses your Apify token; the platform supplies the request identity used for billing.

For an input error (400), correct the request rather than retrying it. For a temporary busy or source error, retry later with backoff. Report beta problems on the Actor's Issues tab with the HTTP status, error code and check time; never include your API token. The standard Apify MCP integration runs the batch Actor; this HTTP endpoint is a separate API interface.

Price a whole list in one run

keywords takes a list — one item or your whole buying list. Each keyword returns its own summary row at $0.02, so 50 keywords = $1.00, answered in one run.

Measured on 2026-08-21 (real run): 50 camera keywords, default settings → all 50 summaries in 110 seconds, $1.00.

Two limits to know before a big list: the run's default time limit is 300 seconds (the 50-keyword run used about a third of it), and the platform stops a run when it reaches its Maximum cost per run setting (default $3.00). Summary-only runs usually reach the time limit before the cost limit; about 130 keyword/status searches fit the measured pace. Representative evidence can reach the cost limit around 75 searches when all 10 rows are returned. Full evidence at the default 120-row sample reaches it after about 11 searches. both counts sold and on-sale as separate searches. Raise both run limits together or split a large list.

Input example

{
    "keywords": ["ポケモンカード リザードン psa10", "Leica M6"],
    "itemStatus": "sold_out",
    "evidenceMode": "representative",
    "maxItemsPerKeyword": 120,
    "convertToUsd": true
}

Japanese keywords give the best coverage. English brand names (Leica, Rolex, Supreme) also work well.

evidenceMode accepts summary, representative, or full. Representative returns up to 10 price-bearing listings chosen at even ranks across the sampled price range; full returns every price-bearing listing sampled. A fresh Console form starts with summary selected. Saved Tasks and API calls that omit evidenceMode use the legacy includeIndividualItems setting: true returns full evidence, while false or an omitted legacy setting returns summary only. An explicit evidenceMode always wins, even when the legacy setting is also supplied.

maxItemsPerKeyword takes 30 to 600 (120 is one result page), and the summary is worked out from the listings actually sampled, not from every sold listing that exists. It changes returned evidence volume only in full mode. itemStatus: "on_sale" looks at listings still for sale instead; days-to-sell is then not returned.

Output

One price_summary row per keyword/status, followed by the evidence rows selected by evidenceMode.

Measured on 2026-08-17 (real run, default input)

{
    "type": "price_summary",
    "keyword": "ポケモンカード リザードン psa10",
    "status": "sold_out",
    "summaryFor": "ポケモンカード リザードン psa10 (sold_out)",
    "keywordStatus": "ok",
    "totalListingsFound": 7247,
    "sampledListings": 120,
    "priceJpy": {
        "min": 1600,
        "p25": 15725,
        "median": 29100,
        "p75": 70700,
        "max": 470900,
        "average": 48955
    },
    "compSpread": 4.5,
    "mixedKeywordLikely": true,
    "medianDaysToSell": 5,
    "checkedAt": "2026-08-17T17:38:37.488155+00:00",
    "priceUsd": {
        "min": 10.05,
        "p25": 98.77,
        "median": 182.78,
        "p75": 444.07,
        "max": 2957.72,
        "average": 307.49
    },
    "exchangeRateJpyUsd": 0.006281
}

min / p25 / median / p75 / max are the lowest price, the two ends of the middle 50% range, the middle price, and the highest. medianDaysToSell is worked out from when a listing was posted and when it last changed, so read it as roughly how fast things move, not an exact sale date. priceUsd and exchangeRateJpyUsd appear only when USD conversion is on (it is on by default).

totalListingsFound is Mercari's own hit count, and it stops at 15000: that value means "15000 or more", not exactly 15000 (measured 2026-08-18 — camera, カメラ, Nikon, guitar and ギター all report 15000). The statistics are always worked out from the listings actually sampled, so a capped count does not change them.

compSpread is how wide the middle 50% is (p75 ÷ p25), and mixedKeywordLikely is true above 4x. One product sits near 1.5–3x; a higher number means the keyword caught several kinds of thing and no single median describes them. Measured 2026-08-18: sold Nikon Z6 spans 13.9x, because mount adapters, guide books and cases share the keyword with the camera bodies. When you see true, add a price range (e.g. priceMinJpy: 50000) or a more exact keyword and run again.

Representative and full evidence rows include the listing title, JPY/USD price, condition, photo, category id, timestamps, and a direct jp.mercari.com URL. evidenceRole marks each of the ten evenly spaced places from the low end to the high end of the sampled prices; full rows use full_sample.

When a keyword comes back empty

keywordStatus says which kind of empty it is, and an empty result is never charged:

keywordStatus	What it means	What to do
ok	Priced listings were sampled	—
not_found	Mercari has nothing under that spelling	Try the Japanese name (camera → カメラ) or another way of writing the model number (α7 IV / ILCE-7M4)
unknown	Listings exist, but none of the sampled ones carried a price	Narrow the keyword, or try the other itemStatus

When the status is not ok, a hint field spells out the same next step in one sentence.

Use with AI agents

Connect your client to the Apify MCP server, then ask it to discover this Actor by jpmarketdata/mercari-japan-price-checker. The standard MCP integration runs the batch Actor; the HTTP beta is a separate interface. Give the agent the exact input and how to interpret the result:

Use jpmarketdata/mercari-japan-price-checker to research Mercari Japan sold prices for WH-1000XM5. Run with keywords: ["WH-1000XM5"], itemStatus: "sold_out", evidenceMode: "summary", maxItemsPerKeyword: 120, priceMinJpy: 12000, priceMaxJpy: 40000 and convertToUsd: false. Report priceJpy.median, p25, p75, sampledListings, mixedKeywordLikely and checkedAt. Describe the result as sampled sold-out listing prices, not verified payments. Do not start a follow-up run unless I ask.

This example incurs at most $0.02 in summary events for one nonempty keyword/status, before trial credits and any applicable Apify platform charges. Repeated runs are billed again.

Fair use & data

• Reads only publicly visible listing data — no login, no personal data beyond what any visitor sees
• Requests are spaced out to keep the load on Mercari negligible
• Not affiliated with Mercari, Inc. Data is provided for market research; verify before large transactions

FAQ

How can I check Mercari Japan sold prices?

Run Mercari Japan Price Summary with a specific model keyword, itemStatus: "sold_out" and evidenceMode: "summary". It returns a median, middle 50% range and sample count from 30–600 sampled listings. Add a price band or representative evidence to inspect mixed products. This is a current search sample of sold-out-labelled listing prices, not a verified transaction ledger.

Are Mercari sold-out prices confirmed sale prices?

No. sold_out is a source search label; results can include trading items, and the displayed price does not independently confirm payment or a completed transaction. Evidence rows retain their actual status. medianDaysToSell estimates the interval between source timestamps, not a verified sale date.

Does this return Mercari price history or every sold listing?

Each run is a fresh, bounded search sample. Save and rerun an Apify Task with fixed settings to build your own observation history; changing samples can move the median. For up to 2,000 raw item rows per keyword, choose the Mercari Japan Listings Scraper — Bulk Item Data.

Can I export Mercari results or call a price API?

Batch runs support Apify API access and JSON/CSV exports. JSON preserves nested priceJpy statistics and optional evidence. The separate HTTP price lookup beta returns one keyword's summary directly; use batch runs for multiple keywords, both or evidence rows.

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).
• Source failures are reported. Read run logs and partial-result warnings; a successful run can contain fewer completed searches than requested. An acquisition failure is not proof of zero market matches.
• 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 private account or payment records are accessed.

More tools by the same author

• Mercari Japan Listings Scraper — Bulk Item Data
• Mercari Japan Underpriced Deals — Under the Sold Price
• Yahoo! Auctions Japan Sold Price Summary & History
• Daikokuya Japan Buyback — What a Shop Actually Paid
• Hifido Japan Used Audio — What It Actually Sold For
• Japan Used Camera & Lens — Dealer Price vs Sold Price
• Used iPhone & Android Prices Japan — Dealer vs Sold
• Japan Resale Demand — Sold Share & Sales per Day

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 Mercari. 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.