About CarCast

Every used car dealer knows the feeling: you buy a truck at auction on Tuesday, and by Friday the market has shifted. The data you had was already stale when you made the decision.

CarCast was built to solve that problem. We use a proprietary AI forecasting engine — the same class of AI that powers weather forecasting and financial prediction — to generate 30 and 60-day price forecasts for 798 used vehicle segments in the US.

Every competitor in this space — vAuto, Black Book, KBB — tells you where prices are. CarCast tells you where prices are going. That's the difference between reacting and planning.

How it works

We ingest real dealer listing data daily, aggregate it into weekly price signals for each segment, and feed 52+ weeks of history into our forecasting model. The model outputs probabilistic forecasts with P10/P50/P90 confidence bands — so you don't just get a single number, you get a range.

From that range, we derive a trend classification — Rising, Stable, or Softening. If the P50 forecast shows prices rising more than 1.5% over 30 days, the segment is classified as Rising. Declining more than 1.5% is Softening. Everything in between is Stable. These classifications are informational analytics, not financial advice or a recommendation to purchase any vehicle.

Who we serve

Our primary audience is the ~53,000 independent used car dealers in the US, along with auto wholesalers, fleet managers, and car flippers. Anyone who needs to make buying or selling decisions on used vehicles and wants data — not gut feel — driving those decisions.

Our commitment

We believe in transparent, honest forecasting. Our models aren't always right — no forecast is. But we show our confidence intervals, we publish our backtesting results, and we let the data speak. No hype. No black boxes. Just numbers.