Your share, your pricing power, and where you are discounted. Every month.

Ask inside Claude or ChatGPT and get the sold-market read back in seconds: what left dealer websites last month, at what asking price against sticker, how fast, and how that compares with every competitor, nationally and state by state.

Free to add, with the Manufacturer Desk reports preselected. Reports from $0.99 each. Prefer the browser? Sign up and run reports on the web.

Since Jan 2021
monthly sold data, every state
National to city
the grain the sold data carries
Asking at exit
the price when a vehicle left the site
Claude + ChatGPT
or run it in your browser

What the Manufacturer Desk does for you

Share against the whole market

The sold data is national first: brand and model volumes are read against the whole market in one pass rather than assembled from fifty state lists, so the denominator is honest. Brand by brand, model by model, new and used, every month since January 2021.

Pricing power against sticker

Every sold-market row carries the asking price at exit against the sticker, so you can see which brands and models hold sticker, which are discounting, and which way that is moving. Advertised discounting, labelled as such.

Regional grain for incentives and allocation

State and city grain with units, asking price, odometer and days to sell on every row shows where a model is scarce and moving fast, and where it is sitting and discounted, so incentive and allocation calls start from evidence.

Per report

Run any report on this page for its listed price with no subscription.

Reports built for automakers

True national share for your brand and every rival from one national read, the change in points on a year ago, and who holds sticker while who discounts, month by month. The first slide of the monthly share deck, with the numbers behind it.

Your hybrid, plug-in, electric and combustion mix month by month against the market, each powertrain priced and timed, and the mix by state. The electrification read allocation and pricing teams share.

Which trims of one model earn their place: units and share, the asking-price band, the gap to sticker, days to sell, and what each trim holds on the used side. The contenting review starts from evidence instead of anecdote.

The whole used and new market month by month, nationally or in one state: units, median asking price, odometer, days to sell and the sticker gap against a year earlier. The denominator every share and pricing conversation starts from.

The used-market pulse for one nameplate: median asking price, days on market, the certified premium and the top configurations, with the sold band and 12-month series beside it. The fastest read on how a model holds up after it leaves the showroom.

Where a nameplate sells cheapest and dearest by state, with the odometer beside every price and the national cheapest and dearest months. The used-market price map beside the new-vehicle Discount & Scarcity Map.

Ask the sold market directly

The monthly sold data behind these reports is also a tool your assistant can query: units, asking-price percentiles, odometer, days to sell and the sticker gap for any make, model, trim, model year, body type or powertrain, nationally, by state or by city, back to January 2021.

See the Sold Market Data tool

Questions it answers

  • "What is each brand's true national share of new-vehicle sales last month?"
  • "In which states is Nissan discounted more than 2 percent below sticker?"
  • "How fast did the 2026 Subaru Outback take over from the 2025 model year?"

Your Manufacturer Desk connector URL

Add this URL as a custom connector in Claude or ChatGPT. The reports on this page come preselected on the approval screen, where you can add or remove any of them.

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

Name it "MarketCheck AI Desk" when prompted. Plain /mcp works too and shows the full catalog.

Questions before you connect

What is the data behind the Manufacturer Desk?
Monthly sold-vehicle statistics inferred from listings leaving US dealer websites: a sale is recorded when a listing leaves the site, and the price kept is the advertised asking price at that moment. It runs monthly since January 2021 at national, state and city grain, for new and used vehicles, and can be ranked by make, model, trim, model year, body type, drivetrain, powertrain, colour, dealership group or seller website.
What can it not tell me?
Transaction prices. Every figure is an asking price at exit, never a negotiated price, so read the sticker gap as advertised discounting. Sales that never pass through a dealer website are outside it too: Tesla and other direct-to-consumer new sales, auction lanes and private-party sales. There is no ZIP, metro or radius grain (city within a state is the finest), no VIN or rooftop grain, and the latest complete month lags the calendar by one to two months.
How do I connect it?
Add the connector to Claude or ChatGPT once with the URL on this page and the Manufacturer Desk reports come preselected on the approval screen. Claude Code, Codex, Cursor, VS Code and Gemini CLI connect from one line with a Desk key from the Getting Started page. Or sign in and run every report in your browser.

Every insight sits on the largest live automotive dataset in the U.S.

We have been continuously crawling, decoding and normalizing every dealer listing in North America since 2014, and every MarketCheck report cites that data directly.

Listings indexed
5.2B
14M active, refreshed daily from 52,000+ dealer websites.
Off-market vehicles
410M
Delisted, shown at last list price. 1,200+ days of history.
VINs decoded
270M
Year, make, model, trim, spec. US and UK.
API calls / day
16.2M
Serving 4,200+ customers in 38 countries.

600+ data points per VIN

OEM-grade specs, options, packages, build sheet, fuel economy, dimensions, safety ratings.

Continuous since 2014

12 years of price history, depreciation curves, regional variation, seasonality patterns.

Vehicles that cleared the market

We track when listings disappear: the closest observable proxy to a transactable price, recorded at last asking price.

What we ingest

  • Dealer rooftops
  • OEM build data
  • Auction sales
  • NHTSA recalls
  • Title brands
  • Trade-in offers
  • Lease residuals
  • Dealer DMS feeds
  • Pricing scrapes · 24h
  • VIN catalog · weekly

Three steps. No code.

  1. Add MarketCheck to Claude or ChatGPT

    Paste one link into Claude or ChatGPT. It takes about 30 seconds.

  2. Add starter credits

    Your balance is real money, not abstract credits. Reports start at $0.99 each, with a $25 starter minimum. Auto top-up keeps you running, and unused balance is fully refundable.

  3. Ask Claude or ChatGPT a question

    "What's this Camry worth?" "Scan my lot for under-priced cars." "Local market for SUVs in 60601?" Your assistant pulls the report, narrates it, and charges your credits.

Free to add. Prefer the browser? Sign up and run reports on the web.