Pay-per-inference for AI agents: How BlockRun and Incarna use Amazon Bedrock AgentCore payments

Amazon Bedrock AgentCore payments provides a managed way for AI agents to pay for services on demand, using the x402 payment protocol and enforcing spending limits at the infrastructure level. Incarna, a company built by SpreadX, integrated AgentCore payments to allow its agents to pay BlockRun for model inference one request at a time. BlockRun is a pay‑as‑you‑go inference router that serves more than 90 models from over 15 providers, quoting and settling each call independently. By using AgentCore payments, Incarna reduced the effort required to add x402 payment support from months to days, completing the integration in three days with about 200 lines of application code. The system creates a payment session that sets a ceiling and expiry, ensuring agents cannot overspend. Each session enforces the limit at the infrastructure layer, and the agent’s own wallet signs the transaction. Across the beta, agents processed over 1,000 payments ranging from $0.001 to $0.05 per call, each settling individually on the Base network. The integration demonstrates how agents can safely and efficiently pay for inference, with BlockRun handling provider selection and delivery, and Incarna providing a persistent identity and wallet for each agent. This approach offers a governed, on‑demand payment flow that can be adopted by other AI developers needing to purchase services as agents run.