RippleX Lead Says x402 Payments on XRP Ledger (XRP) Near 12 Million

AI Agents Pay for Services On-Chain

Automated purchases by artificial intelligence agents have carried x402 payments on the XRP Ledger to just under 12 million transactions. Agents are buying computing capacity, data and, increasingly, financial services, paying per call rather than through subscriptions. The t54 activity dashboard, which records x402 settlement on the ledger, listed 11,965,947 payments as of Saturday, October 3, with a seven-day trading volume average of 525,618 per day. J. Ayo Akinyele, engineering lead at RippleX, Ripple's blockchain development unit, marked the 10 million milestone in a post on X on Thursday, October 1, noting that volume had climbed from 1 million to 10 million payments in under three months. His post put the recent run rate at about 500,000 payments per day, and he wrote that watching agents autonomously discover, pay for and use services scale this quickly is remarkable, adding that the ledger is increasingly becoming infrastructure for the machine economy. x402 itself is an internet payment protocol that lets software pay for online resources before access is granted. The name comes from the HTTP 402 Payment Required status code, the response that tells a client a resource must be paid for first. Purchases so far concentrate on data and on AI inference, the use of trained models to produce answers or other outputs, and the service range is widening into financial data, risk and credit. These figures are tracked on-chain and sit apart from the XRP price, which settlement activity on the ledger does not directly set. Akinyele frames the growth as evidence that blockchain settlement is taking on a broader role, moving the ledger beyond simple value transfer toward Web3 infrastructure for automated commerce. The milestone lands in a stretch where Ripple has highlighted both accelerating institutional use and growing XRP adoption.

@ja_akinyele · X post

10 million milestone.

View on X

How x402 Settlement Works

Under the x402 flow, a provider selling information or compute to an automated buyer first returns a price and a payment demand instead of the resource itself. The buyer signs a transaction, a facilitator verifies it and submits it to the ledger, and the provider releases the requested data or capacity once settlement confirms. Ripple describes three-to-five-second settlement and predictable transaction costs as the properties that let a program confirm a transfer succeeded before it continues its task. The facilitator model matters because it moves trust off the buyer: the coordinator checks the signed payment before anything is broadcast, so a provider never hands over data against an unverified promise. To connect autonomous software to this flow, Ripple ships an AI starter kit that handles wallet creation, balance queries and payment tracking, so an application can monitor spending and confirm that a task actually completed. The kit also supports ongoing balance checks, meaning an agent's operator can cap exposure while the agent keeps running. Purchases can be denominated in XRP or in RLUSD, Ripple's dollar-backed stablecoin, which lets both sides of a transaction price services in dollars rather than in a volatile asset. Pricing in RLUSD also gives agents a stable unit of account when budgets, spending caps and repayment schedules are set in dollar terms. Akinyele attributes most current activity to data purchases and AI inference, while his latest update points to services in financial data, risk and credit, with feeds of the kind a blockchain oracle delivers. The 10 million milestone followed the 1 million mark he discussed in July, and the gap between the two, less than three months, is the clearest measure of how quickly agent demand has compounded. Wider commercial use is already being scoped: t54 co-founder Chandler Fang outlined a framework on Friday, August 21 in which agents operate inside user-defined spending limits, with budgets, merchant restrictions and approval requirements for delegated purchases, plus potential credit for software buyers that can build a reliable identity and repayment record.

From Data Buys to Machine Credit

The most load-bearing record here is the milestone post itself: Akinyele, writing for RippleX's engineering side, fixed the seven-day run rate at 500,000 payments per day and tied the growth directly to agents that discover, pay for and use services on their own. That primary document, not secondhand commentary, anchors the path from 1 million in July to 10 million by October 1. COINOTAG's read is that the next leg is financial: the service list already runs to financial data, risk and credit, and Fang's August framework sketches agents with budgets, merchant limits and repayment histories, the raw material for underwriting. Dashboards currently run through October 3; anything later is unreported.