Guides are open. Infrastructure is your choice.
You can read this site and use its examples without creating an account. MultichainAPI.com does not sell API plans, issue keys, meter calls, or operate a hosted endpoint. Any node or provider you choose has its own pricing and terms.
Build a workload worksheet
Start with expected users, actions per user, and requests per action. Then account for refreshes, pagination, backfills, and retries. Keep assumptions visible so the estimate can change with the feature.
planned reads = users × actions per user × reads per action
Illustration only:
100 users × 12 visits × 4 reads = 4,800 reads
Add refreshes, retries, and background work separately.Compare more than a base price
| Check | Question to ask |
|---|---|
| Metering | Are all methods counted the same way? |
| History | Is the historical depth you need included? |
| Request limits | Which burst, sustained, and concurrent limits apply? |
| Data products | Are indexed datasets, streams, or exports priced separately? |
| Overages | Does the service stop, throttle, or bill extra? |
| Operations | What access controls and support terms are documented? |
Use evidence instead of a fictional price card
Check the selected provider’s current plan documentation, run a representative workload, and record the assumptions behind your estimate. Do not treat the arithmetic above as a quote, usage forecast, or statement of any provider’s metering rules.
The provider evaluation guide turns these questions into a repeatable decision process. The retry guide helps account for the additional work that failed requests can create.