Decision model pricing comparison — the full catalog, live

The whole point of this page

Decision-model pricing is scattered across vendor docs. This gateway keeps the on-sale catalog in one place: the Model Square and every model page render the full catalog with per-million input prices, vendors and release dates — synced from the live /v1/models feed on every deploy. Output tokens are free on every model.

For the current table, open the Model Square — it pairs the catalog with search and vendor filters, and every row links to a per-model page.

How billing works here

Cost = round(input tokens × input price/M) + round(output tokens × output price/M). With output priced at $0 catalog-wide, you pay for what you send. Balances are kept in integer quota (1,000,000,000 quota = $1) so ledgers carry no floating-point drift.