Solutions

Charge for AI usage per request.

Charge for inference, tools, and API calls with credit-backed settlement. Customers pay as they go, and supported collateral can keep earning yield.

Charge per call

Charge each request using signed guarantees instead of invoices.

Yield on collateral

Deposits route through Aave and accrue yield while they back payments.

TypeScript and Python SDKs

Use TypeScript or Python clients with your existing HTTP stack.

Use cases

Built for ai companies workflows

Give each stakeholder the payment primitives they need: instant authorization, clear limits, and settlement that stays easy to audit.

For API teams

Monetize each inference

Price model calls, tool calls, embeddings, and data lookups without forcing prepaid balances.

Meter premium endpoints

Add payment checks to routes that are too costly to leave unmetered.

Let customers start sooner

Let users and agents pay from an existing wallet without creating a prepaid account.

For finance and ops

Reconcile by settlement cycle

Group many small requests into clear receivables and on-chain settlement records.

Earn on supported collateral

Supported collateral can earn yield while it backs payments.

Process

Start with ai companies in 3 steps

The path is deliberately small: configure the payment session, verify each request, then settle the cycle with durable records.

  1. 01

    Add pricing to protected routes

    Define which API endpoints require payment and what each request costs.

  2. 02

    Verify payment headers

    Use the middleware to validate x402-compatible guarantees before compute is spent.

  3. 03

    Collect settled usage

    Receive batched, auditable settlement instead of chasing invoices for tiny charges.

FAQ

Common questions about ai companies

Can 4Mica charge for individual API calls?

Yes. Each protected request can carry a signed payment guarantee, so you can meter inference, tools, data, and other API usage per call.

Do customers need a new account balance inside my app?

No. Customers can pay through credit-backed x402 flows instead of maintaining a separate prepaid balance with you.

How does this affect expensive AI workloads?

Payment verification happens before the response, so you can avoid serving costly compute to requests without a valid guarantee.

Something else on your mind?Contact us

Ready to build for ai companies?