Estimate cost per request, scenario, and month.
Estimate token-driven API cost with current public rate cards, shareable scenarios, and model-specific source links.
Model cost planning without the gimmicks
Use public model pricing, token estimates, and shareable URL state to pressure-test a workflow before it becomes a real budget line.
This is approximate, but it is useful for turning a real prompt or model output into a practical planning input.
Turn a rate card into a workload estimate
Start with the model, then estimate prompt size, response size, any cached context, and how often the request runs. The calculator separates per-request cost from monthly volume so you can see whether a rate difference is trivial or operationally meaningful.
Use the provider pages when you are still deciding between ecosystems, then use this calculator when you already have a model shortlist.
What this estimate does not know
- Provider discounts, enterprise contracts, or free-tier credits not shown on the public rate card.
- Latency, rate limits, retries, tool-call overhead, or orchestration costs outside the token bill.
- Future pricing changes after the page’s listed verification date.