Agentic payment activity on Coinbase’s Base network has surpassed 100 million transactions, indicating that machine-to-machine payments are moving beyond the proof-of-concept stage in onchain environments.
According to a new Chainalysis report, wallets that interacted with Coinbase’s x402 protocol generated more than 100 million transactions on Base within about nine months of launch.
The x402 protocol allows software agents to make onchain payments directly via web requests. When an agent requests access to a resource, such as a data feed or API, it can automatically complete a stablecoin payment without human permission.
Much of x402’s early growth was driven by a memecoin experiment called PING, which required users to make a payment through the protocol in order to mint tokens. The project attracted large numbers of users looking to acquire the token, triggering a massive increase in transaction activity.
Although activity decreased after the PING frenzy subsided, usage did not decline. According to Chainalysis, transaction volumes have stabilized while the value of transfers has increased.
At the beginning of 2025, transactions worth more than $1 accounted for approximately 49% of the total value transferred via x402. By early 2026, that figure had risen to 95%, indicating that the protocol goes beyond just micropayments.
Cumulative agentic transfer volumes on Base. Source: Chain analysis
Onchain data points to growing use of agentic payments
The rise of AI tools has led to renewed interest in agent payments. Proponents say crypto networks are well-suited for these transactions because they can move money and process payments automatically 24 hours a day, without requiring a user to approve each purchase.
Several crypto industry leaders, including Coinbase CEO Brian Armstrong and Circle CEO Jeremy Allaire, have argued that AI agents could soon account for a significant portion of onchain activity. Former Binance CEO Changpeng Zhao has expressed a similar view, describing cryptocurrency as the “native currency” of AI agents.
Early versions of machine-to-machine payments in crypto already exist. Decentralized computer networks allow users and applications to pay for GPU resources on demand, while decentralized data marketplaces allow applications to purchase data sets and blockchain information through automated transactions.
Weekly wallet retention for agentic payments on Base is trending upward. Source: Chain analysis
Interest in the concept extends beyond crypto. A recent Forrester report highlights Stripe’s Machine Payments Protocol as a potential catalyst for reviving micropayments through AI agents.
Analysts at Bernstein said AI agents could drive demand for stablecoins, which are well suited for frequent, low-value payments, which Coinbase’s x402 protocol highlights.
Magazine: AI-driven hacks could kill DeFi – unless projects take action now