Custom Model Pricing Calculator

Use current vendor, reseller, private deployment, or negotiated pricing instead of relying on a preset. Enter custom input, output, and cached-input rates per million tokens, then apply your real token volume and request frequency.

Calculate with explicit assumptions

Model price cards change and enterprise agreements can differ from public rates. Custom pricing keeps the calculation auditable: every projection is derived from the values entered in the browser. It is useful for newly released models, NVIDIA NIM endpoints, internal inference services, regional pricing, and contract-specific discounts.

Review model pricing before committing budget

Provider prices, cache rules, batch discounts, context tiers, and regional availability can change. Treat the result as a planning estimate, verify the selected rates against the current provider documentation, and use custom pricing when your contract differs from the public list price.

Related calculator modes