Model intelligence

Historical multi-step cost calculator.

This earlier tool combines user-supplied call counts, token use, and assumed outcomes. It does not measure quality or verify savings. For current model pricing research, use the API cost calculator.

Find model or provider

Showing up to 200 matching options per step. Search the full catalog by name, provider, or model ID. Estimates use prices observed within seven days with no unresolved price review. Historical rates remain on model pages.

Workflow runs / monthAssumed acceptable outcomes (%)

AI step 1Model / provider

Input tokens / callVisible output tokens / callAdditional reasoning tokens / callCache-read percentageCalls including retries / run

models.dev · retrieved 2026-11-01 · Price history

Reasoning tokens are additional to visible output and use the sourced reasoning rate, or the output rate when no separate rate is reported. Cache estimates require a sourced cache-read rate. Excludes tax, volume tiers, cache writes, external tools, and platform fees. Provider billing terms may differ. Formula alm-cost-2.0.