Blockchain analytics platform Chainalysis has reported that wallets interacting with Coinbase’s x402 protocol have generated more than 100 million transactions on Base since the payment standard launched, offering one of the clearest signs yet that machine-to-machine payments are beginning to gain traction on public blockchains.
Built to let software agents pay for online resources automatically, x402 allows a stablecoin payment to be completed as part of a web request. Under the model, an AI agent seeking access to an API, data feed, or digital service can settle the payment directly without requiring a user to manually approve each transaction.
A large portion of the protocol’s early activity came from PING, a memecoin project that required users to make a payment through x402 to mint tokens. According to Chainalysis, the token launch attracted significant user participation and produced a sharp increase in transaction counts across the network.
While activity cooled after interest in PING faded, Chainalysis said usage levels remained well above pre-launch levels. The report noted that transaction volumes have stabilized over time even as larger-value transfers account for a growing share of activity.
Data cited by Chainalysis showed that payments exceeding $1 represented about 49% of the value transferred through x402 in early 2025.
By early 2026, those transactions accounted for roughly 95% of value moved through the protocol, a change that the blockchain analytics firm said suggests usage is extending beyond small experimental payments.
Coinbase expands infrastructure for AI-powered transactions
Recent developments across Coinbase’s ecosystem have placed x402 at the center of several projects focused on AI-driven payments and automation.
In May, Coinbase-backed Base introduced Base MCP, a framework that connects Base Accounts to AI assistants such as ChatGPT, Claude, Codex, and Cursor through the Model Context Protocol. Base said the system allows users to manage transfers, swaps, balance checks, transaction reviews, and x402 payments through conversational interfaces while keeping final approval with the account holder.
At the same time, Base integrated a range of decentralized finance applications, including Moonwell, Morpho, Uniswap, Avantis, Aerodrome, and Virtuals, enabling AI agents to assist users with lending, liquidity management, token swaps, and derivatives activity. According to Base, every transaction still requires user confirmation through a separate approval window.
Coinbase has also been building supporting infrastructure around x402 through Agentic.market, a platform designed to help autonomous agents discover and purchase services using $USDC. Earlier figures cited by Base showed x402 activity had already reached roughly 165 million settled transactions involving more than 480,000 agents.
Outside Coinbase’s ecosystem, interest in machine payments continues to spread. Amazon Web Services added support for Coinbase’s x402 standard through Amazon Bedrock AgentCore Payments, allowing AI agents to purchase services using $USDC. Stripe has also introduced support for x402 on Base, while AllUnity has adopted the standard for its own agentic payment layer.
Growing attention around these systems comes as industry executives increasingly argue that AI agents could become important users of blockchain networks.
Coinbase Chief Executive Officer Brian Armstrong has said AI agents could account for a meaningful share of future onchain activity, while Circle Chief Executive Officer Jeremy Allaire has made similar comments regarding autonomous payments. Former Binance Chief Executive Officer Changpeng Zhao has described cryptocurrency as the “native currency” of AI agents.
Separate research has also pointed to growing commercial interest in the sector. A Forrester report identified Stripe’s Machine Payments Protocol as a potential catalyst for AI-driven micropayments, while Bernstein analysts highlighted Coinbase’s x402 protocol and argued that increasing use of AI agents could support demand for stablecoins because of their suitability for frequent, low-value transactions.
The trend has emerged alongside rapid growth forecasts for agentic AI. According to Fortune Business Insights, the agentic AI market could reach $139.19 billion by 2034, while MarketsandMarkets estimated that agentic AI software services could expand at a 46.8% compound annual growth rate through 2032.