MarketCheck
Market Data ToolsBeta

Sold Market Data

Monthly sold-vehicle statistics from tens of thousands of US dealer websites (the Market Insights tool).

from $3.00 per call, priced by request shape

What it returns

Raw data for your assistant to read.

Ranked rows of monthly sold-vehicle statistics, as a list of columns and one row per month, geography and group you ask for: the units sold, the priced sample, the median and the mean advertised asking price at the time of sale, the median odometer and the average days to sell, with a thin flag on small samples. Any other statistic (the Low end and High end percentiles, the 95th, the revenue proxy, days on the selling site, the gap to sticker) comes back when you rank or filter by it. Rank by make, model, trim, model year, body type, drivetrain, transmission, powertrain, colour, dealership group or selling website, nationally, by state or by city, for any window of whole months back to January 2021. Every answer echoes the request it applied and the warnings that shaped it, and states the units and dollars it cost.

Use it when

  • You want the exact-car band: what a 2022 Camry SE was listed at when it sold in Texas last month, on a priced sample you can cite.
  • You want a true national ranking or share, not fifty state lists stitched together.
  • You want a 12-month trend with the odometer beside the price: a retention curve, a seasonality read, an EV share series.
  • You want who sells, ranked: selling websites in your cities, or dealership groups against each other.

Not when a report already answers it

  • You want one make and model read as a finished, shareable report: Brand & Model Market Pulse, or Market Movers for what is clearing faster than local supply.
  • You want individual cars: this tool returns statistics, never a VIN. Use the recently sold search for rows.
  • You want this week: the data is monthly and the latest complete month lags the calendar by one to two months.

Pay per call

Priced by the shape of the question.

$3.00 per call unit

Debited from the prepaid balance after a successful call: a standard question is 1 unit ($3.00) and the largest shape is 16 units ($48.00). No subscription: a call costs the largest shape it uses plus any add-on, each answer states the units and dollars it cost, and a refused or failed call costs nothing.

Request shapePrice

A standard question: national or state rows for up to 12 months, or city rows for one month, ranked by up to two vehicle dimensions or by make, model and body type

For example: The top 15 used models in Texas last month (1 unit, $3.00)

$3.00 1 unit

Pinned to one or more named selling websites (source as a filter)

For example: CarMax used sales in Texas, by model (2 units, $6.00)

$6.00 2 units

A wide shape: city rows over several months or above a 5,000-row limit, or three or more ranking dimensions that go beyond make, model and body type

For example: The top makes in every Texas city, month by month for six months (3 units, $9.00)

$9.00 3 units

Dealership groups, as a filter or a ranking dimension

For example: The top 20 dealership groups nationally (5 units, $15.00)

$15.00 5 units

Selling websites ranked inside one state or its cities

For example: The top selling websites in Dallas (5 units, $15.00)

$15.00 5 units

Selling websites ranked nationally, or with no state named

For example: The top 25 selling websites in the country (10 units, $30.00)

$30.00 10 units

A selling-website ranking that also splits by make or model

For example: The top selling websites in Texas, by make (10 units, $30.00)

+$15.00 +5 units

A window that starts more than 24 months back

For example: The top makes nationally, Jan to Jun 2023 (2 units, $6.00)

+$3.00 +1 unit

Real results by desk

The prompt, what the assistant sent, what came back.

Every excerpt is a real result, trimmed to the fields that matter, with its date. Every price in it is an advertised asking price.

Dealer

The top 15 used models in Texas last month

"Show the top 15 used models sold in Texas last month with the median asking price and days to sell."

What the assistant sends

{
  "inventory_type": "used",
  "summary_by": "state",
  "state": "TX",
  "ranking_dimensions": "make,model",
  "top_n": 15
}

1 unit: $3.00

What came back (Data month Aug 2026, taken Oct 16, 2026), trimmed

{
  "applied": { "summary_by": "state", "date_window_source": "default", "inventory_type": "used", "state": ["TX"] },
  "rows_total": 15,
  "columns": ["month", "make", "model", "rank", "sold_count", "price.count", "price.p50", "miles.p50", "dom_active.mean", "thin"],
  "rows": [
    ["Aug 2026", "Ford", "F-150", 1, 6892, 5916, 30988, 85942, 38.48, false],
    ["Aug 2026", "Chevrolet", "Silverado 1500", 2, 5594, 4740, 31062, 78938, 37.22, false],
    ["Aug 2026", "RAM", "Ram 1500 Pickup", 3, 3584, 3153, 30974, 78138.5, 37.96, false],
    ["Aug 2026", "Toyota", "Camry", 4, 3504, 3127, 24715, 59320, 32.04, false],
    ["Aug 2026", "Nissan", "Rogue", 5, 3343, 3091, 20627, 48999, 37.38, false],
    ["Aug 2026", "Toyota", "RAV4", 7, 2602, 2427, 27998, 51983, 34.1, false],
    ["Aug 2026", "Nissan", "Altima", 10, 2473, 2107, 18972, 62014, 46.58, false]
  ]
}

What to do with it: Three pickups lead Texas on volume at a $31,000 median, but the Camry turned six days faster at $24,715 and the Altima nine days slower at $18,972: the buy list is the fast turners inside the band you stock.

Which websites sold the most used cars in Dallas

"Which dealer websites sold the most used cars in Dallas last month?"

What the assistant sends

{
  "inventory_type": "used",
  "summary_by": "city_state",
  "state": "TX",
  "city": "Dallas",
  "ranking_dimensions": "source",
  "top_n": 10,
  "date_from": "2026-08-01",
  "date_to": "2026-08-31"
}

5 units: $15.00

What came back (Data month Aug 2026, taken Oct 16, 2026), trimmed

{
  "rows_total": 10,
  "columns": ["month", "city", "source", "rank", "sold_count", "price.p50", "miles.p50", "dom_active.mean"],
  "rows": [
    ["Aug 2026", "Dallas", "sewelllexus-dallas.com", 1, 410, 44052, 27886, 40.24],
    ["Aug 2026", "Dallas", "drivecasa.com", 2, 404, 18370, 111469.5, 11.93],
    ["Aug 2026", "Dallas", "toyotaofdallas.com", 3, 349, 28913, 50560, 23.56],
    ["Aug 2026", "Dallas", "miniofdallas.com", 4, 219, 35990, 31207, 54],
    ["Aug 2026", "Dallas", "dallasleasereturns.com", 7, 193, 33113, 52407, 18.21],
    ["Aug 2026", "Dallas", "drivetime.com", 9, 157, 20795, 74119, 25.45]
  ]
}

What to do with it: The second-biggest used seller in Dallas turns 111,000-mile cars in 12 days at an $18,370 median: the conquest target for a sub-$20k store.

Dealer Group

New-vehicle sales by brand, nationally, with days to sell

"Rank new-vehicle sales by brand nationally last month, with the median asking price and days to sell."

What the assistant sends

{
  "inventory_type": "new",
  "summary_by": "national",
  "ranking_dimensions": "make",
  "top_n": 10,
  "date_from": "2026-08-01",
  "date_to": "2026-08-31"
}

1 unit: $3.00

What came back (Data month Aug 2026, taken Oct 16, 2026), trimmed

{
  "rows_total": 10,
  "columns": ["month", "make", "rank", "sold_count", "price.count", "price.p50", "dom_active.mean"],
  "rows": [
    ["Aug 2026", "Toyota", 1, 281630, 275512, 43532, 32.41],
    ["Aug 2026", "Ford", 2, 141999, 135604, 51990, 95.66],
    ["Aug 2026", "Chevrolet", 3, 124102, 122695, 46306, 65.82],
    ["Aug 2026", "Honda", 4, 113107, 109845, 37444, 44.12],
    ["Aug 2026", "Kia", 5, 76753, 76080, 35046, 68.14],
    ["Aug 2026", "Hyundai", 6, 70896, 70095, 35649, 98.46],
    ["Aug 2026", "Nissan", 7, 62729, 58911, 32939, 88.7],
    ["Aug 2026", "Lexus", 10, 37041, 36718, 62587.5, 36.08]
  ]
}

What to do with it: Toyota sold twice the new volume of Ford and turned it in 32 days against 96, and Hyundai and Nissan sat about three months too: brand turn summed correctly nationally in one call, before a group weighs its next rooftop.

The top 20 dealership groups nationally

"Rank the top 20 dealership groups in the country by used units sold last month, with price and days to sell."

What the assistant sends

{
  "inventory_type": "used",
  "summary_by": "national",
  "ranking_dimensions": "dealership_group_name",
  "top_n": 20,
  "date_from": "2026-08-01",
  "date_to": "2026-08-31"
}

5 units: $15.00

What came back (Data month Aug 2026, taken Oct 16, 2026), trimmed

{
  "warnings": [ { "code": "sparse_dimension_rows_excluded", "message": "Rows with no value for dealership_group_name are excluded from this result, and that dimension is documented as populated on ~38% of rows. sold_count here counts only rows that have a value for every dimension you asked for, so it covers a fraction of sales rather than the whole market." } ],
  "rows_total": 20,
  "columns": ["month", "dealership_group_name", "rank", "sold_count", "price.p50", "miles.p50", "dom_active.mean"],
  "rows": [
    ["Aug 2026", "Carmax", 1, 74805, 24998, 39617, 32.97],
    ["Aug 2026", "Carvana", 2, 73484, 25590, 47297, 28.16],
    ["Aug 2026", "Lithia Motors Inc.", 3, 28175, 24795, 50771, 44.15],
    ["Aug 2026", "AutoNation Inc.", 4, 25991, 25778, 50437, 39.61],
    ["Aug 2026", "Asbury Automotive Group", 5, 13041, 28279, 46438, 30.83],
    ["Aug 2026", "Group 1 Automotive Inc.", 6, 13029, 27998, 48734.5, 34.9],
    ["Aug 2026", "Penske Automotive Group Inc.", 7, 11463, 31790, 32905, 42.82],
    ["Aug 2026", "Berkshire Hathaway Automotive", 8, 11215, 25900, 46010, 26.53]
  ]
}

What to do with it: CarMax and Carvana each outsell the next group nearly three to one and Carvana turns nearly five days faster; the warning says the table covers sales with a known group, not every sale.

Claims

How a 2020 Explorer held its asking price over 12 months

"How has the 2020 Ford Explorer held its value over the last year, with the odometer beside the price?"

What the assistant sends

{
  "inventory_type": "used",
  "summary_by": "national",
  "make": "Ford",
  "model": "Explorer",
  "year": "2020",
  "ranking_dimensions": "none",
  "date_from": "2025-09-01",
  "date_to": "2026-08-31"
}

1 unit: $3.00

What came back (Data month Aug 2026, taken Oct 16, 2026), trimmed

{
  "rows_total": 12,
  "columns": ["month", "sold_count", "price.count", "price.p50", "miles.p50", "dom_active.mean"],
  "rows": [
    ["Aug 2026", 2466, 2226, 22582.5, 83234, 41.18],
    ["Jun 2026", 2335, 2099, 22995, 80726.5, 38.45],
    ["Mar 2026", 2707, 2465, 22687, 78806, 43.15],
    ["Dec 2025", 2442, 2218, 22971.5, 76481, 45.14],
    ["Sep 2025", 2538, 2310, 24137, 73094, 41.11]
  ]
}

What to do with it: Down 6.4 percent in a year while the median odometer rose 10,000 miles, on 2,300 to 2,700 sales a month: a cohort-honest curve to read a loss month against, not a mix-blended average.

Auction

Where a 2021 CR-V listed highest when it sold, state by state

"Where did a used 2021 Honda CR-V list highest when it sold last month? Every state with at least 20 sales."

What the assistant sends

{
  "inventory_type": "used",
  "summary_by": "state",
  "make": "Honda",
  "model": "CR-V",
  "year": "2021",
  "ranking_dimensions": "none",
  "min_sold_count": 20,
  "date_from": "2026-08-01",
  "date_to": "2026-08-31"
}

1 unit: $3.00

What came back (Data month Aug 2026, taken Oct 16, 2026), trimmed

{
  "applied": { "min_sold_count": 20, "ranking_ignored": true },
  "rows_total": 24,
  "columns": ["month", "state", "sold_count", "price.count", "price.p50", "miles.p50", "thin"],
  "rows": [
    ["Aug 2026", "VA", 52, 48, 26994, 61325, false],
    ["Aug 2026", "TX", 79, 76, 23998, 70667, false],
    ["Aug 2026", "OK", 20, 20, 28289, 42345.5, true],
    ["Aug 2026", "IL", 36, 36, 26983.5, 55578, false],
    ["Aug 2026", "FL", 105, 94, 24516, 55231, false],
    ["Aug 2026", "CA", 103, 100, 25349.5, 56987, false]
  ]
}

What to do with it: A $3,000 spread between Texas and Virginia on the same model year, with the odometer beside each median; the Oklahoma row carries thin: true on 20 priced sales, so it is direction, not a price to ship against.

Researcher

The EV share of used sales, month by month

"How is the EV share of used sales moving, month by month, over the last year?"

What the assistant sends

{
  "inventory_type": "used",
  "summary_by": "national",
  "ranking_dimensions": "fuel_type_category",
  "date_from": "2025-09-01",
  "date_to": "2026-08-31"
}

1 unit: $3.00

What came back (Data month Aug 2026, taken Oct 16, 2026), trimmed

{
  "rows_total": 60,
  "columns": ["month", "fuel_type_category", "rank", "sold_count", "price.p50", "miles.p50", "dom_active.mean"],
  "rows": [
    ["Aug 2026", "ice", 1, 1616067, 23590, 60379, 45.98],
    ["Aug 2026", "ev", 2, 49082, 30897, 27276.5, 35.96],
    ["Aug 2026", "hybrid", 3, 35612, 25512.5, 55918.5, 45.25],
    ["Apr 2026", "ev", 2, 59352, 28910, 28387, 43]
  ]
}

What to do with it: Battery-electric cars were 2.9 percent of used sales in Aug 2026 and cleared ten days faster than combustion; the five rows a month are the numerator and the denominator of a share chart.

Shopper

What a 2022 Camry SE was listed at when it sold in Texas

"What was the median asking price of 2022 Toyota Camry SE models that sold in Texas last month?"

What the assistant sends

{
  "inventory_type": "used",
  "summary_by": "state",
  "state": "TX",
  "make": "Toyota",
  "model": "Camry",
  "year": "2022",
  "trim": "SE",
  "ranking_dimensions": "none",
  "date_from": "2026-08-01",
  "date_to": "2026-08-31"
}

1 unit: $3.00

What came back (Data month Aug 2026, taken Oct 16, 2026), trimmed

{
  "rows_total": 1,
  "columns": ["month", "inventory_type", "state", "sold_count", "price.count", "price.mean", "price.p50", "miles.p50", "dom_active.mean", "thin"],
  "rows": [
    ["Aug 2026", "used", "TX", 92, 78, 22205.12, 22094, 77757, 22.92, false]
  ]
}

What to do with it: A $22,094 median on 78 priced sales, with the mean at $22,205 and a 77,757-mile median odometer: the figure an offer, or a settlement, should start from. Ask for price.p25 and price.p75 to get the band around it.

Limits and traps

Read before you lean on a number.

  • inventory_type is required and every enum is lowercase: used, new, certified or all. A typo in a filter value matches nothing rather than erroring, and the call is still charged, so read the applied block the answer echoes.
  • Rows with fewer than 30 priced sales carry thin: true. Read them as direction, not a price.
  • A call covers whole months, up to 12 at a time, from January 2021 to the latest complete month. A longer window, or one that starts before January 2021, is refused before it is sent and costs nothing; a window that runs past the latest month is cut at it.
  • The result carries the median and the mean asking price. To get a percentile or another statistic back, rank by it or pass it as a filter such as ">=0".
  • The service returns up to 2,000 rows unless the call sets limit (at most 20,000), and the result holds as many as fit in about 60,000 characters, plus rows_total; narrow with filters or a lower top_n rather than paging.
  • A dealership group is a sparse dimension (populated on roughly 38 percent of rows), and the answer warns when rows without a group were excluded.
  • inventory_type=all pools new and used into one row: only sold_count, price.sum and the .count fields describe a real population there. Add inventory_type to ranking_dimensions for separate rows.
  • Sales are inferred from listings that left dealer websites; every price is the advertised asking price at the time of sale, never a transaction price. US dealers only, so the tool is not offered to a UK account.

For developers

The names your code needs.

MCP tool

get_market_insights

Scope (one permission per tool)

data:get-market-insights

Endpoint

POST https://ai.marketcheck.com/mcp

Call it like every other tool, from a connected assistant or with an agent's OAuth client-credentials token; the result arrives as JSON in content[0].text. See the tools reference and server-side access.

Try it in your assistant

Connect once, then paste a prompt from this page.

Claude connects in one click, and ChatGPT (Plus or higher) adds MarketCheck as a custom plugin from the connector URL; Cursor, VS Code, Gemini and other MCP clients use the same URL. The Market Data Tools are selected by default when you approve access.

Questions

The short answers.

What is a call unit?

The step the price moves in: $3.00. A standard question (national or state rows for up to 12 months, ranked by vehicle dimensions such as make, model or trim) is one unit. City rows over several months, dealership groups and selling-website rankings cost more units, as the table above shows, and every answer states the units and dollars it cost.

What counts as a sale?

A listing that left a US dealer website and met the sold-detection rules. Prices are the advertised asking prices at the time the vehicle left, not negotiated transaction prices. Coverage is US franchise and independent dealers, monthly since January 2021.

Is there a subscription?

No. Every call is paid from the prepaid balance the reports use, only when it succeeds. A refused or failed call costs nothing.

Can I download the whole data set?

No. Every answer is a ranked summary: the service returns up to 2,000 rows unless the call sets a limit (at most 20,000), and the result your assistant reads holds as many as fit in about 60,000 characters, with the total beside them.

What if my AI agent has no account?

It can pay per call with a Stripe Shared Payment Token, at the same price for the same request shape.

Reports on the same data

When you want the finished read.

Brand & Model Market Pulse

Used-market pulse: pricing, CPO premium, days-on-market, top configurations.

$0.99 per run

Market Movers

What to stock next: nameplates clearing faster than local supply.

$4.99 per run

Full Market Capture

The mother of all market snapshots: your whole market, captured in one run.

$9.99 per 10k listings processed

The other data tools

Same connector, same balance.

Active listings

$0.01 per call

Recently sold

$0.02 per call

VIN history

$0.02 per call

NeoVIN decode

$0.16 per call

Price prediction

$0.18 per call