Question 1

How does the AI spend calculator estimate cost?

Accepted Answer

It multiplies your chosen model's per-token pricing by your entered call volume and average prompt/completion token counts, entirely in your browser — no data is sent to a server. It then compares that cost against cheaper models in the same family and estimates savings from prompt caching and batch routing.

Question 2

How much AI spend is typically recoverable as waste?

Accepted Answer

Most teams find 28–40% of their AI spend is recoverable, typically from three sources: using a frontier model for a task a cheaper model handles equally well (token maxing), resending a large static system prompt on every call without caching it (context tax), and retries or agent loops that never should have run.

Question 3

Does the calculator require a signup or my real usage data?

Accepted Answer

No. The calculator runs entirely client-side with no network calls and no signup. It only estimates cost from the model, call volume, and token counts you enter — it does not need or request your actual usage data.

Stop guessing your AI spend.