Japan Land Price API — Official Appraisals & 40-Year History

Every official land-price survey site in Japan, with what it is worth today and what it was worth every year back to 1983.

{
    "address": "東京都 中央区銀座4丁目2番4",
    "prefecture": "東京都",
    "municipality": "中央区",
    "municipalityCode": "13102",
    "latitude": 35.671989,
    "longitude": 139.765335,
    "year": 2025,
    "pricePerSqm": 60500000,
    "pricePerTsubo": 200000090,
    "yearOnYearChangePercent": 8.6,
    "change10YearPercent": 79.0,
    "changeSincePeakPercent": 0,
    "zoning": "商業",
    "floorAreaRatioPercent": 800,
    "nearestStation": "銀座",
    "distanceToStationMeters": 100,
    "priceHistory": [{ "year": 2002, "pricePerSqm": 14100000 }, "…"]
}

No scraping. No API keys. No rate limits. Every answer comes from a registry compiled from Japanese government open data and shipped inside the Actor.

---

What is in it

46,988 survey sites	25,563 from 地価公示 + 21,425 from 都道府県地価調査
1,933 municipalities	every prefecture, city, ward, town and village with a survey site
43 years of prices	the appraised price for each site for every year from 1983
26 attributes per site	price, change, area, zoning, FAR, coverage ratio, current use, building structure, nearest station and distance, front road and its width, and a written description of the surrounding area

Japan runs two official appraisals a year and this Actor carries both:

• 地価公示 (koji) — appraised as of 1 January by the national government under 地価公示法. This is the number quoted when the news says "land prices rose x%".
• 都道府県地価調査 (chosa) — appraised as of 1 July by each prefecture. It covers areas outside the 公示 grid, so together they give roughly twice the coverage of either alone.

---

Why this is hard to get otherwise

The source is published as 国土数値情報 Shapefile/GeoJSON archives with numbered columns (L01_008, L01_051, …) whose meaning is not in the file. The published attribute spec does not line up with what actually ships: 標準地コード is three columns in the data but two in the spec, and the order around 駅名 / 交通施設との近接状況 differs as well. Guess wrong by one and you publish the wrong price against the wrong site.

Every column in this Actor was pinned by profiling the real value distributions across thousands of features — direction fields hold the eight compass points, 用途地域 holds the twelve zoning codes, 建蔽率 lands in 0–80, 容積率 in 0–800, and so on. The mapping is regression-tested: the most expensive site in the country must come out as 銀座4丁目2番4 at ¥60,500,000/m², which is the published 2025 figure.

---

Use cases

• Value a Japanese property. Find the survey sites within a kilometre, take the price per square metre, and you have the appraisal basis every Japanese valuer starts from.
• Check what can be built on a site. zoning, floorAreaRatioPercent, buildingCoverageRatioPercent, cityPlanningArea and frontRoadWidthMeters are the inputs to a Japanese development feasibility calculation.
• Screen a market before you fly. Sort a prefecture by yearOnYearChangePercent to find where prices are actually moving, or by changeSincePeakPercent to find what is still below its bubble-era high.
• Build a time series. 43 years of appraisals per site, already cleaned of the zeros that mark "not yet surveyed".
• Give an AI agent a Japanese property tool. It runs as an MCP tool: an agent can ask "what is land worth near this station, and is it rising?" and get typed rows back.

---

Input

{
    "prefectures": ["東京都"],
    "municipalityCodes": ["13102"],
    "keyword": "銀座",
    "near": { "latitude": 35.6717, "longitude": 139.765, "radiusMeters": 1000 },
    "surveyTypes": ["koji"],
    "zoning": ["商業"],
    "minPricePerSqm": 1000000,
    "sortBy": "price",
    "maxResults": 100
}

Every field is optional; an empty input returns the most expensive sites in Tokyo.

Field	Meaning
prefectures	Prefecture names, with or without the suffix (東京 = 東京都)
municipalityCodes	5-digit 全国地方公共団体コード — narrower and faster than a prefecture
keyword	Substring over site name, address, municipality, nearest station and area description
near	{latitude, longitude, radiusMeters} — results come back sorted by distance
surveyTypes	koji (地価公示) and/or chosa (都道府県地価調査)
zoning	1低専, 1中専, 1住居, 近商, 商業, 準工, 工業, 工専, …
currentUse	住宅, 店舗, 事務所, 工場, 駐車場, 田, 畑, …
minPricePerSqm / maxPricePerSqm	Price band in JPY per square metre
includePriceHistory	Set to false for a flat, smaller output
sortBy	price, priceAscending, yearOnYearChange, distance
maxResults	Hard cap on rows — this is what determines the cost of the run

Output

One row per survey site. Notable fields:

Field	Description
pricePerSqm / pricePerTsubo	Appraised land value, JPY. Tsubo is the unit Japanese agents quote
yearOnYearChangePercent	Official change against last year's appraisal
change5YearPercent, change10YearPercent	Computed from this site's own history
changeSincePeakPercent	Against the site's highest appraisal ever — negative means still below its bubble peak
zoning, buildingCoverageRatioPercent, floorAreaRatioPercent, cityPlanningArea	What may be built
nearestStation, distanceToStationMeters	As recorded by the appraiser
frontRoad, frontRoadDirection, frontRoadWidthMeters	Road access, which drives value in Japan
surroundings	The appraiser's written description of the neighbourhood
priceHistory	[{year, pricePerSqm}], oldest first, zeros removed
municipalityCode	The join key for every Japanese government statistic
distanceMeters	Radius searches only

---

Pricing

Pay per event: a small charge to start the run, plus a charge per site returned. There are no proxy costs, because there is nothing to proxy.

---

Data sources

Everything is public, free and downloaded from the publisher. There is no scraping anywhere in this Actor, at build time or at run time.

Source	Publisher	Licence
国土数値情報 地価公示 (L01-25)	国土交通省	公共データ利用規約 第1.0版 (PDL1.0)
国土数値情報 都道府県地価調査 (L02-24)	国土交通省	公共データ利用規約 第1.0版 (PDL1.0)
郵便番号データ	日本郵便	Japan Post asserts no copyright over the postal code data

PDL1.0 permits commercial use and redistribution and requires the source credit plus a statement that the data was edited. Reproduced here and in every run's RUN_SUMMARY:

「国土数値情報（地価公示）」（国土交通省）（https://nlftp.mlit.go.jp/ksj/）をもとに jp-data-engine が加工して作成

「国土数値情報（都道府県地価調査）」（国土交通省）（https://nlftp.mlit.go.jp/ksj/）をもとに jp-data-engine が加工して作成

郵便番号データ（日本郵便株式会社）をもとに jp-data-engine が加工して作成

The registry is recompiled from the publishers' current releases every time the Actor is rebuilt, and each row carries the datasetBuiltAt date it was answered from.

---

Known limits

Stated up front rather than discovered later:

• These are survey points, not every parcel. Japan has ~47,000 official survey sites, not a price for every plot. The standard practice is to value a property from the nearby sites, which is what the radius search is for.
• 地価公示 is as of 1 January, 地価調査 as of 1 July, and the latest editions here are 2025 and 2024 respectively. Mixing the two in one ranking mixes two appraisal dates.
• Zoning is blank for about 9,000 rural sites — they sit outside a 用途地域, which the source records as empty rather than as a category.
• Price history starts when the site was first surveyed, which for many sites is well after 1983. priceHistory is only as long as the site is old.
• This is not a transaction database. These are appraisals, not actual sale prices.

---

Local development

npm install
npm run build:data   # downloads the government data and compiles the registry into data/
npm test             # 17 tests, including the field-mapping regression
npx apify run        # run with the default input

Part of the jp-data-engine Japan data suite — Japanese data, structured for everyone else. See also Japan Address Normalizer, Geocoder & Romanizer for turning a Japanese address into coordinates and a municipality code you can feed straight into this Actor.