How it works
One VIN. One clear decision.
What actually happens between typing 17 characters and reading a verdict — including the parts where we decline to give you a number.
You enter a VIN
One required field. VINLytix uppercases as you type, strips spaces and punctuation people paste in from listings, counts characters, and validates the ISO 3779 check digit — a checksum over the other sixteen characters that catches most typos before anything is queried.
We validate before we query
Length, forbidden characters, and the checksum. The letters I, O and Q never appear in a VIN, so if you have typed one the report tells you which and what it probably should be, rather than returning an empty result.
We identify the vehicle
The VIN goes to NHTSA vPIC, the official US federal database of manufacturer-submitted vehicle specifications. That returns year, make, model, trim, body, engine, fuel, drive type, transmission, manufacturer and assembly plant.
We check safety recalls
Using the decoded year, make and model, we query NHTSA's published recall campaigns and return the defect description, the safety consequence and the manufacturer's remedy for each.
We analyse what you gave us
Mileage against the expected figure for the vehicle's age. Asking price against market value, where a valuation is available. Advertised specification against the decoded specification, if you supplied the listing details.
We score, and we say how much we knew
A Deal Score out of 100, and separately a Data Coverage percentage. Then a verdict, the red flags, the two-column list of what was and was not checked, and a pre-purchase checklist adapted to the findings.
The two numbers, and why there are two
The Deal Score answers “how attractive does this look?” It combines price against market, mileage against age, recall status and vehicle age, each with a published weight. Components with no data are dropped from the calculation rather than scored zero, so a missing input never looks like a bad result — and the score is always shown with a confidence label reflecting how much of the total weight was available.
Data Coverage answers a different question: “how much could you actually check?” It scores a fixed eight-domain checklist, so it cannot be inflated by quietly shortening the list. Two of those domains — recall repair status for an individual VIN, and full accident and title history — are currently unreachable for every report, which caps coverage at 75%. That cap is deliberate and visible.
Showing only the first number would let a confident-looking 88 hide the fact that more than half the relevant record was never examined. The full methodology is published.
What we will not do
- Invent a market value. With no licensed pricing provider connected, the report says valuation is unavailable. It does not produce an approximate number from the VIN.
- Call a vehicle clean. The strongest phrasing anywhere in a report is “no issue was found in the data that was checked”.
- Imply an accident history was checked. It was not. Every report says so in its own section.
- Estimate salvage or parts value. It depends on damage, component condition, title status, local demand and scrap prices. None of that is in a VIN.
- Claim the Deal Score is scientifically validated. It is a presentation heuristic for comparing listings, and it says so.
Frequently asked questions
Do I need an account?
No. There is no signup, no email gate and no account of any kind. Enter a VIN and read the report.
How long does a report take?
A few seconds. You will see each step as it runs, because the progress list is driven by the work actually happening on the server rather than a scripted animation.
Why does the report show steps I did not expect, or skip ones I did?
Only steps that actually run are shown. If you did not enter mileage, there is no mileage step. If no valuation provider is configured, there is no valuation step. A progress bar that counts through work nobody did would be its own small dishonesty.
Can I change my mind about whether I'm buying or selling?
Yes, at any time, and it does not re-run anything. The perspective switch at the top of the report re-reads the same data for your side of the transaction. VINLytix never asks which one you are before showing results.