Blockchain infrastructure company Alchemy has launched a system that allows autonomous AI agents to purchase compute credits and access blockchain data services using onchain and USDC on Base wallets.
According to the company announcementthe first version enables AI agents to directly query blockchain networks, verify ownership of non-fungible tokens (NFTs), view wallet balances across multiple chains, and access live token price data while planning additional networks and services.
If an agent exhausts its prepaid compute credits, Alchemy issues a payment request that can be automatically settled in USDC (USDC) on the database, allowing the agent to continue operating without human intervention.
The company says agents can fund accounts with as little as $1 in USDC, and once the funds are credited, continue making API calls until the balance is depleted and another automatic payment is required.
The system uses Coinbase’s x402 payment standard to convert a “402 Payment Required” HTTP response into an automatic billing trigger. x402 is an open standard that allows web services to request onchain payments directly via HTTP responses, enabling machine-to-machine transactions without manual invoicing.
Nikil Viswanathan, CEO of Alchemy, told Cointelegraph that the system is aimed at developers creating autonomous decentralized finance (DeFi) agents, portfolio management bots and other multi-step online workflows.
He said several major crypto apps, including Robinhood Crypto, Uniswap, OpenSea, Aave and 0x, already rely on Alchemy to power transactions, adding:
Now AI agents can access the same infrastructure autonomously, without a human touching it. This is where agent economics gets its own set of keys.
Related: The sports intellectual property industry cannot defend itself against artificial intelligence without blockchain
Cryptocurrency companies and developers are accelerating agent deployment
Over the past year, AI agents, i.e. software systems that can independently make decisions and perform tasks based on pre-defined goals and real-time data, have gained increasing attention. Almost a quarter (23%) of organizations examined by McKinsey in November said it was expanding its utilize of agent systems.
On February 6, AI platform AI.com, founded by Crypto.com CEO Kris Marszalek, said it plans to launch an autonomous AI agent for retail users, capable of conducting stock trades, automating workflows and handling routine digital tasks.
A few days later, Coinbase introduced “Agentic Wallets,” a crypto wallet infrastructure designed to enable AI agents to autonomously spend, earn, and trade digital assets, including executing onchain transactions such as managing DeFi positions, rebalancing wallets, and paying for services.
Meanwhile, the Monad developer community announced the results of this week’s Moltiverse Hackathon, during which 16 winners were selected from a number of “agent + token” projects.
The winners included a programmable venture capital agent that rates and invests automatic multiplayer battle controlled by artificial intelligence arenasAI dating network where agents “represent their people” and a trading card sports competitions managed by software, not people.
Warehouse: Crypto loves Clawdbot/Moltbot, Uber ratings for AI agents: AI Eye
