Other catalogs and data services answer what exists right now. WhatModel answers what's changing, what changed, and what's coming next.
Every price, context, and capability is recorded with a timestamp.
Other catalogs and data services show you the present. We show you the past. Watch a model's input price fall 40% over a quarter, or see when a provider quietly doubled a context window. Nobody else carries this memory.
Each price cut, context bump, new capability, or deprecation is an event.
Not a row in a table — an object with a timestamp, a payload, a permalink, and a ticker spot. The market doesn't have a state, it has a stream.
One model, every source — side by side.
The same model is listed in multiple places at different prices. We aggregate sources and surface the deltas so you pick the cheapest route, not the loudest one.
Watchlists, alerts, and a portfolio of models that's yours.
Other catalogs and data services are stateless. We remember which models matter to you, ping you when their pricing or capabilities change, and surface your slice of the market on every page.
Every cell, badge, and chart point is an entry point.
Press ⌘K to jump anywhere. Click a provider chip to filter. Click a capability to pivot the catalog. Add models to a persistent compare tray from any surface. It feels like Bloomberg, not a docs page.
The WM Score is a 0–100 blended rating that combines third-party quality benchmarks with publicly observable operational signals — price, context window, capability breadth, uptime, and provider redundancy. Missing signals are dropped from the denominator rather than substituted, so the score is honest about what it does and doesn't know.
Weights are tuned periodically as the market shifts. The full formula and current weights are proprietary.
