What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Estimate Anthropic API usage by pricing each token category separately—uncached input, cache writes, cache reads, and output—then add any feature charges or pricing modifiers. Only after calculating usage should you account for prepaid credits or a sales-arranged invoice: credits affect how a charge is paid, not the token rates used to calculate it.
Calculate the cost of a cached API request
Use the selected model’s current input and output rates, and count tokens according to how they are billed. A cached request may contain four distinct token categories: ordinary input, cache creation, cache reads, and generated output. Applying the ordinary input rate to all input tokens will misstate the charge.
As an Amazon Associate I earn from qualifying purchases.
For a model using Anthropic’s standard cache multipliers, the estimate is:
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchcost = (uncached_input_tokens × I + 5m_cache_write_tokens × 1.25I + 1h_cache_write_tokens × 2I + cache_read_tokens × 0.1I + output_tokens × O) / 1,000,000
Here, I is the model’s input price per million tokens and O is its output price per million tokens. Check the pricing page for the current model rates and cache rules; the usual 0.1× cache-read multiplier has model-specific exceptions. Include any separately priced server-side feature and applicable pricing modifier. This is an estimate of usage charges, not an account invoice calculator.
Price each token category
- Uncached input: token count multiplied by the model’s base input rate.
- 5-minute cache creation: at the standard rate, token count multiplied by 1.25 times base input.
- 1-hour cache creation: at the standard rate, token count multiplied by 2 times base input.
- Cache reads: generally token count multiplied by 0.1 times base input, subject to model-specific exceptions.
- Output: generated-token count multiplied by the model’s output rate.
Decide whether a cache write is worthwhile
Anthropic’s published multipliers imply that, before other costs, one cache read can repay the 5-minute write premium and two reads can repay the 1-hour write premium. That break-even rule assumes comparable token pricing and that the same cached content is actually reused within its lifetime. Model-specific rates, cache duration, and reuse patterns determine the real result.
Turn request estimates into a monthly forecast
- Choose the model and copy its current input and output rates and applicable cache multipliers from Anthropic’s pricing page.
- For representative requests, estimate uncached input, each cache-write duration, cache reads, output, and any separately charged features.
- Multiply each category’s estimated charge by the expected monthly request volume. Allow for variation in prompt size, generated output, cache reuse, and feature use rather than assuming every request is identical.
- Keep the resulting usage estimate separate from credits or invoice payments. Check the organization’s Console billing settings for the funding arrangement and available balance.
Understand credits, invoices, and paid Claude plans
“Monthly credits” can refer to different arrangements. Console API usage is generally paid from prepaid credits, while some organizations have monthly invoicing arranged through Anthropic Sales. Neither should be confused with optional usage credits described for certain paid Claude plans.
| Arrangement | How it works | What to verify |
|---|---|---|
| Console API and Playground prepaid credits | Most organizations buy credits before use; usage is charged against them at current pricing. Purchased credits are available immediately and expire one year after purchase. If the balance is exhausted, API calls and Playground use stop until more credits are added. | Check the Console Billing page for available balance, credit expiration, usage, and spending controls. Anthropic Help Center. |
| Sales-arranged monthly invoicing | Eligible organizations with an arrangement established through Anthropic Sales are billed monthly in arrears for aggregate usage at standard pay-as-you-go prices. This is invoicing, not a recurring credit grant. | Confirm whether monthly invoicing is configured for your organization in Console or with your account contact. Anthropic Help Center. |
| Usage credits for certain paid Claude plans | Optional usage credits may be available after included plan limits. They are distinct from included limits and are billed at standard API rates; availability and expiration depend on the applicable terms and, in some cases, jurisdiction. | Check the relevant Claude account billing details and terms. Do not assume these credits are the same as Console prepaid credits. Anthropic Help Center. |
Public pricing information cannot establish an individual organization’s credit balance, monthly allowance, eligibility, or next expiration date. Those are account-specific details, so check the relevant Console or Claude account Billing and Usage settings rather than inferring an allowance from a subscription name.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Reconcile an estimate against actual usage
For actual API usage, Anthropic’s usage-report endpoint supports breakdowns including uncached input, 5-minute cache creation, 1-hour cache creation, and cache reads. Compare those reported token categories with your estimate, then separately compare usage charges with Console credits or invoice lines. The Console Billing page shows organization credit balance and usage.
Anthropic’s pricing documentation also gives a worked Managed Agents example: 50,000 input tokens at $5 per million, 15,000 output tokens at $25 per million, and one hour of Managed Agents runtime at $0.08 total $0.705. Counting 40,000 input tokens as cache reads at the stated rate changes that example total to $0.525. It includes Managed Agents runtime, so it is not a bare Messages API request price. See the pricing documentation.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
Recommended Free Tools
