Decision tools

AI Usage Receipt Calculator

An AI usage receipt reprices disjoint token counts using an explicit rate card. Uncached input, cache reads, cache writes and billed output are counted separately. Import complete provider usage objects locally or enter aggregate counts, then choose your rates. The result is a token-cost scenario, not proof of an actual payment, a subscription saving or equivalent performance from a different model. Shared links and PNGs contain aggregate assumptions rather than raw prompts or file contents.

Open the live AI Usage Receipt calculator - Token-cost receipt →

Example scenario

The synthetic example contains 1,000 uncached input tokens, 8,000 cache-read tokens, 2,000 cache-write tokens and 1,000 output tokens. Hypothetical per-million rates of 2, 0.20, 2.50 and 8 dollars produce subtotals of 0.002, 0.0016, 0.005 and 0.008 dollars, totaling 0.0166 dollars. These rates are illustrative, not a provider quote.

What the inputs mean

Calls represented: Enter your own value in count. The initial value is illustrative; replace it with measured usage or a documented assumption.
Uncached input tokens: Enter your own value in count. The initial value is illustrative; replace it with measured usage or a documented assumption.
Cache-read input tokens: Enter your own value in count. The initial value is illustrative; replace it with measured usage or a documented assumption.
Cache-write input tokens: Enter your own value in count. The initial value is illustrative; replace it with measured usage or a documented assumption.
Total billed output tokens: Enter your own value in count. The initial value is illustrative; replace it with measured usage or a documented assumption.
Uncached input rate: Enter your own value in USD / 1M tokens. The initial value is illustrative; replace it with measured usage or a documented assumption.
Cache-read rate: Enter your own value in USD / 1M tokens. The initial value is illustrative; replace it with measured usage or a documented assumption.
Cache-write rate: Enter your own value in USD / 1M tokens. The initial value is illustrative; replace it with measured usage or a documented assumption.
Output rate: Enter your own value in USD / 1M tokens. The initial value is illustrative; replace it with measured usage or a documented assumption.

What the result means

The receipt exposes exactly which counts and rates produced each subtotal. A mixed-model import with one rate card is a counterfactual repricing scenario, not the models’ combined paid bill.

Assumptions

This reprices usage with one editable, flat rate card. It is not an invoice, subscription saving claim or provider quote. Use separate receipts for different models, tiers or time periods. Tool, media, taxes and negotiated charges are excluded. Output includes billable reasoning where the usage format includes it.
Default inputs are original illustrative examples, not provider quotes or measured customer outcomes.
Shared links preserve numeric assumptions and engine version; downloads do not include raw imported files.

Where the prices come from

Rates in this tool are editable user assumptions. No provider API is called and no input is refreshed automatically. Verify current billing units and terms separately before using the scenario for a purchasing decision.

Formula and methodology

Repriced token cost = (uncached tokens x uncached rate + cache-read tokens x cache-read rate + cache-write tokens x cache-write rate + billed output tokens x output rate) / 1,000,000. The importer uses the existing provider-usage normalizer and validates complete numeric counters before aggregation. Cumulative counters, invalid JSONL and conflicting streaming updates are rejected. Use a separate receipt when a model, tier, rate or billing period changes.

Interpretation guide

Compare alternatives with the same workload assumptions.
Stress-test output-heavy, retry-heavy, cache-miss, and power-user cases before committing budget.
Verify source links and production logs before using the estimate for billing decisions.

Limitations before production billing decisions

Treat ByteCosts calculations as planning estimates, not final billing totals. Real invoices can differ because token mix, retry rate, cache hit rate, rate limits, taxes, gateway fees, regional pricing, and negotiated discounts change the effective cost.

Verify the provider source before production billing decisions, then compare the estimate with your own logs or invoice once production traffic is live.

AI Usage Receipt Calculator. ByteCosts. https://bytecosts.com/tools/ai-usage-receipt/