Coinbase x402 Update Enables Usage-Based AI Payments

Coinbase x402 Update Enables Usage-Based AI Payments

Coinbase has introduced a usage-based payment model for its x402 protocol, targeting AI-driven services with variable costs. The change removes a key limitation that restricted payments to fixed-price transactions.

The update, announced Thursday by the Coinbase Developer Platform, introduces the “Upto” scheme for agentic tasks such as data queries and large language model inference. The system allows buyers to approve a maximum spend while sellers define upper price bounds, with final costs calculated after execution. The protocol operates on an Ethereum Virtual Machine (EVM) framework and supports ERC-20 tokens.

Can Usage-Based Payments Unlock Agentic Commerce Scale?

Previously, x402 only supported exact, fixed-price payments, limiting its utility for services where compute demand fluctuates. This created inefficiencies, with users either overpaying for simple queries or underpaying for complex workloads. The new structure aligns pricing with actual resource consumption, a model already standard in cloud computing but less common in onchain payment systems.

Still, the shift reflects broader industry alignment toward agentic commerce, where autonomous systems execute tasks and settle payments in real time. Compared to traditional API billing systems, which often rely on offchain reconciliation, onchain usage-based settlement could reduce latency and improve transparency in high-frequency machine interactions.

“Until now, x402 only supported exact, fixed-price payments,” Coinbase Developer Platform stated. “But it blocked an entire category of services where the cost depends on usage.”

The integration of the CDP Facilitator also enables gasless transactions, lowering friction for frequent, small-value payments.

Governance of the protocol has also shifted, with Coinbase transferring control to the Linux Foundation. The x402 Foundation now includes participation from major technology firms such as Google, Microsoft, and Amazon Web Services, signaling cross-industry interest in standardized payment rails for AI workloads.

Adoption will depend on whether developers integrate x402 into production-grade agent systems, with early usage metrics and enterprise partnerships likely to serve as the next catalyst.

Read more