Japan Used Truck Prices — By Mileage, Not Just Year

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

What it does: Pick a truck maker and model and get what a used one costs on Truck Bank Japan today, split by how far it has run. You enter: A maker and a model — for example ISUZU and ELF. Years are optional. You get: The typical vehicle price and range in yen; the same for each mileage band, with how many trucks stand in it; how many listings publish a price at all; the year mix; the prefectures. Read from 5 to 12 of the site's own pages. Price: $0.02 per model. No results = no charge, and a model whose every listing says "ask the dealer" is not charged either. +$0.002 per truck if you also want the listings themselves, one row each (off by default).

Example: enter ISUZU / ELF → 1,105 of 1,786 listings publish a price · typical ¥3,080,000 · range ¥268,000–¥12,580,000 · under 200,000 km: 823 trucks, typical ¥3,400,000 · 200,000–500,000 km: 267 trucks, ¥2,398,000 · 500,000 km and over: 13 trucks, ¥1,958,000 (real run, 2026-09-10)

Unofficial — not affiliated with Truck Bank. Reads public pages only.

Why the split by distance run

Truck Bank publishes a price guide of its own, and it is a grid of year against price band. It has no distance axis at all. Yet distance run is what a used truck is priced on, and the site's own search filters on it — so the split exists in its data and nowhere in its published figures.

That is the whole of this Actor's claim: the year table is something the site already gives away, and the mileage table is not.

Measured on 2026-09-10, ISUZU ELF (real run):

Distance run	Trucks	Cheapest	Typical
Up to 200,000 km	823	¥268,000	¥3,400,000
200,000–500,000 km	267	¥790,000	¥2,398,000
500,000 km and over	13	¥968,000	¥1,958,000

How the figures are taken

Truck Bank prints the matching count on every results page, accepts any row offset, and can order the results by vehicle price. So the truck standing at position N/4 of that order is the lower end of the middle 50%: reading it costs one page instead of the 56 pages a full sweep of that model would take.

Every figure in priceJpy is therefore one real listing read at its own place in the line — not an average over a sample. priceBasis says ladder when the figures were read that way and exact when the whole model fitted on one page and was read entire.

The one way that can break is checked on every page. The ordering is a search parameter, and this site ignores a value it does not recognise without saying so — measured 2026-09-10, both an unknown sort value and no sort at all answer with the site's newest-first default. A position read out of an unordered list would be a real price in the wrong column, so every page read is checked for climbing prices, the five published figures are checked against each other, and the run fails rather than publishing them. sortCheck in the record says it happened.

What "price" means here

Nearly half of Truck Bank's stock carries no figure at all — it says "ask the dealer". Every statistic here is built on the listings that publish a price, and the record says how big the silent half was: totalTrucks counts every listing, trucksWithPublishedPrice counts the ones behind the figures, publishedPriceRate is the share. On the measured ELF search that was 1,105 of 1,786.

The main figures are the vehicle price. The total payable is published on only some listings, so it comes with totalPayableCoverage saying how many showed one, and it is taken over the rows the run read rather than off the ladder.

Output

One row per model. Measured on 2026-09-10 (real run, maker ISUZU / model ELF), trimmed:

{
  "type": "used_truck_price_summary",
  "maker": "ISUZU", "model": "ELF",
  "makerOnSite": "いすゞ", "modelOnSite": "エルフ",
  "status": "ok",
  "totalTrucks": 1786,
  "trucksWithPublishedPrice": 1105,
  "publishedPriceRate": 0.6187,
  "trucksRead": 173,
  "priceJpy": {"min": 268000, "p25": 2200000, "median": 3080000,
               "p75": 4378000, "max": 12580000},
  "priceBasis": "ladder",
  "byMileageBand": [
    {"band": "Up to 200,000 km", "fromKm": null, "toKm": 200000,
     "trucks": 823, "minPriceJpy": 268000, "medianPriceJpy": 3400000,
     "maxPriceJpy": null, "trucksRead": 40, "priceBasis": "ladder"},
    {"band": "200,000–500,000 km", "fromKm": 200000, "toKm": 500000,
     "trucks": 267, "minPriceJpy": 790000, "medianPriceJpy": 2398000,
     "maxPriceJpy": null, "trucksRead": 40, "priceBasis": "ladder"},
    {"band": "500,000 km and over", "fromKm": 500000, "toKm": null,
     "trucks": 13, "minPriceJpy": 968000, "medianPriceJpy": 1958000,
     "maxPriceJpy": 5390000, "trucksRead": 13, "priceBasis": "exact"}
  ],
  "medianMileageBand": "Up to 200,000 km",
  "mileageKm": {"min": 1000, "p25": 67500, "median": 150000,
                "p75": 251500, "max": 1171000},
  "mileageBasis": "sample",
  "trucksWithAnImpossibleDistance": 2,
  "totalPayableJpy": {"min": 979910, "p25": 1638440, "median": 2640940,
                      "p75": 4430000, "max": 12780000},
  "totalPayableCoverage": 0.7457,
  "byYear": [{"year": 2026, "trucksRead": 1, "medianPriceJpy": 9757000}],
  "dealerPrefectures": [{"prefecture": "千葉県", "trucks": 27}],
  "countCheck": {"verdict": "mileage_bands_do_not_add_up",
                 "trucksCountedInThePage": 1105,
                 "trucksCountedInThePageDescription": 1105,
                 "countSources": "agree",
                 "trucksAcrossMileageBands": 1103,
                 "trucksInNoMileageBand": 2,
                 "mileageBands": "do_not_add_up"},
  "filterCheck": {"verdict": "narrowed",
                  "shareOfRowsPrintingTheModelAskedFor": 1.0,
                  "shareOfRowsCarryingAPublishedPrice": 1.0,
                  "shareOfBandRowsInsideTheirBand": 1.0},
  "sortCheck": {"verdict": "monotonic", "pagesChecked": 5,
                "pagesOutOfOrder": 0},
  "requestsUsed": 11,
  "fetchedAt": "2026-09-10T08:33:56+00:00"
}

The three columns that are taken over the rows the run read rather than from the site's own counts say so: mileageBasis, byYearBasis and totalPayableBasis all read sample. The band counts, the model counts and the price figures do not — they are the site's own numbers and its own listings.

countCheck is Truck Bank's arithmetic, checked. The site prints its count twice, in the page and in the page description, and the mileage bands ought to add back up to the model. Both hold most of the time and neither is quietly repaired: the ELF run above found 1,103 trucks across the three bands against 1,105 in the model — two listings that fall in no band — and said so rather than dropping the difference.

One more thing the site says and this Actor does not repeat: two ELF listings that day printed 129,482,000 km and 99,417,000 km, which is a dealer's typing rather than a truck. They stay in the band the site put them in and in the listing rows, and they are kept out of mileageKm, where one of them alone would print a range nobody could use. trucksWithAnImpossibleDistance counts them.

Individual trucks

Turn on the listings and a second kind of row follows, one per truck: maker, model, year, mileageKm, priceJpy, totalPayableJpy, bodyForm, inspection, dealerPrefecture, truckId and a link to the site's own page. These are the listings already read for the summary, so they add rows and not waiting. At most 220 of them: eleven pages of twenty, less whatever a truck read twice takes off.

No dealer name, no inquiry number and no phone number is returned, on either kind of row. Every seller on Truck Bank is a dealer rather than a private person and the numbers on the page are call-forwarding lines, but none of it is read off the page at all, so the prefecture the truck stands in is the only thing here about who is selling it.

What makes a run fail

A run that cannot answer honestly stops instead of returning an empty file, and says which of three things happened: the site did not answer, the site refused the page, or the site answered with something these figures cannot be built from. That last one covers a page with no count on it, a page that counts trucks and sends none, results that are not in price order, results that are not the model asked for, listings with no price arriving on a search for priced ones, mileage bands that came back unfiltered, and a model code the site has reassigned to a different truck.

A model with nothing listed is not a failure — it is an answer, it comes back with status: not_found and a note about what to try, and it is not charged. Nor is a model whose every listing says "ask the dealer": that comes back as no_published_prices with the count of trucks that are there, and it is not charged either.

Load discipline

One page every 2.0 seconds, at most 12 pages a whole run, no login, no cookie, no browser and no proxy. Neither number can be reached from the input page. Individual truck pages are never opened — everything priced here is on the results page already — and nothing reads a member page, a saved search or a dealer's contact form.

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

• Carsensor Japan Used Car Prices & Mileage by Model
• Mercari Japan Sold Prices — What Items Really Sell For
• Yahoo! Auctions Japan Sold Prices — Median, Range, Bids

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 Truck Bank. 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.