API monetization: pricing models and per-request payments

API monetization is how an API provider earns revenue from the requests it serves. Common models include subscriptions, usage-based billing, prepaid credits, and payment per request. AI agents add a new kind of customer that can pay at request time. Explore agentic payments on Tempo.

An API that returns valuable data or computation has a cost for every response. Pricing turns that usage into revenue. The right model depends on who calls the API, how often, and how much each call is worth to them.

What is API monetization?

Direct monetization charges for access to the API itself. Customers pay for a plan, for the usage they consume, or for each request.

Indirect monetization uses the API to grow another business. A payments platform, for example, earns on the transactions its API enables rather than on the API calls.

This guide focuses on direct models and how to collect payment for them.

What are the main API pricing models?

Model	How it works	Fits	Tradeoff
Freemium	Free tier with paid upgrades	Developer adoption	Free usage carries cost
Subscription tiers	Monthly plan with a usage allowance	Predictable, steady usage	Light users overpay, heavy users hit limits
Usage-based billing	Metered usage, invoiced after the period	Variable usage by known customers	Collection risk and invoicing work
Prepaid credits	Customers buy a balance that usage draws down	Teams that want spending control	Balances to manage and top up
Per-request payment	Each call is paid at request time	Occasional callers and AI agents	Needs a low-cost way to pay small amounts

Many providers combine models. A subscription can serve existing customers, while per-request payment serves callers who will never sign up for an account.

Why do AI agents change API monetization?

Agents find and call APIs while they work on a task. That breaks several assumptions in traditional API sales:

• No account setup. An agent cannot be expected to create an account, enter card details, and copy an API key for each new provider.
• Small, bursty usage. An agent can make a handful of calls to one API and never return. A monthly plan is a poor fit.
• Price checks at runtime. The agent needs to read the price and decide within its budget before it pays.

Per-request payment solves these problems if paying is cheap and machine-readable. Read agentic commerce for the broader context.

How do per-request payments work with MPP?

The Machine Payments Protocol (MPP) is an open protocol co-authored by Tempo and Stripe. It uses HTTP to request and verify payment:

1. A client calls a paid endpoint.
2. The API responds with HTTP 402 Payment Required and its payment terms.
3. The client authorizes payment from a compatible wallet.
4. The API verifies payment and returns the response with a receipt.

For high-frequency usage, a session can let charges accumulate and settle periodically instead of making an onchain payment for every request. AI payment processing with MPP covers one-time payments, sessions, and subscriptions in more detail.

What should you build before charging per request?

Metering. Count the unit you charge for accurately, and make it visible to customers.

Pricing and access rules. Decide which endpoints are paid, what each costs, and whether any callers get different terms.

Receipts and records. Return a receipt with each paid response and store the payment reference with the request, so finance can reconcile revenue. See payment reconciliation.

Abuse controls. Payment does not replace rate limits, input validation, or monitoring.

Refund policy. State how failed or disputed requests are handled. Stablecoin payments do not include card-network chargeback rules, so define the process in your terms.

Where Tempo fits

Tempo is the blockchain built for payments. It supports MPP so APIs can charge per request and receive stablecoin payments that settle near-instantly with sub-cent network fees. Pay-as-you-go sessions support metered usage, and transfer memos connect payments to your records.

You keep control of prices and access rules. Start with the agents documentation, or see agentic payments on Tempo.

FAQ

What is API monetization?

API monetization is how a provider earns revenue from its API. Direct models charge for access through subscriptions, usage-based billing, prepaid credits, or payment per request. Indirect models use the API to grow another product, such as a platform that charges for transactions its API enables.

How do you monetize an API?

Choose a pricing unit that matches the value customers get, such as requests, records, or compute time. Meter usage accurately, publish clear prices, and pick a collection method: invoices for subscriptions and metered plans, prepaid balances, or payment with each request. Many providers combine models for different customer types.

What is a pay-per-call API?

A pay-per-call API charges for each request rather than a monthly plan. The caller pays the listed price at request time, and the provider verifies payment before returning the response. This model suits occasional users and AI agents that do not have an account with the provider.

Can AI agents pay for API calls?

Yes. With the Machine Payments Protocol (MPP), an API can respond with HTTP 402 Payment Required and its price. An agent with a compatible wallet authorizes payment within its spending limits, and the API verifies it before returning the response.

Can an API charge less than one cent per request?

On Tempo, yes. Tempo supports sub-cent payments, and pay-as-you-go sessions can combine many small charges into periodic settlement. Some funding and custody providers process only whole cents, so check those constraints before setting sub-cent prices.