Blockchain analytics firm Chainalysis said on 14 October 2026 that more than 138 million x402 transactions had occurred on Base, the Coinbase-incubated blockchain, and that over 87% of x402 payments were worth between a hundredth of a cent and a dime.
Chainalysis calls Coinbase the builder of x402, a protocol that, Fortune reports, lets bots pay for online services in crypto without a person approving each purchase. Chainalysis’s own post on X put cumulative x402 transactions on Base at 58.8k in Q3 2025 and 138.5M in Q3 2026.
Business Pill · USAGE-BASED PRICING
A one-minute explainer of usage-based pricing: paying for work done rather than for seats. It teaches the general idea only and says nothing about any organisation in this story.
The key insight: As we read it, the unit of this market is the API call, not the subscription: Chainalysis says the most common x402 payment is between one and ten cents, an amount it says a card network cannot process economically.
Chainalysis studied x402 in its 2026 North America report, calling it “one of the most popular digital payments rails that lets AI agents spend money”. It describes Base as “the most popular blockchain for agentic commerce”.
Fortune, which reported the findings on 14 October, said cumulative x402 transactions on Base rose from about 1,500 in mid-2025 to more than 138 million by the third quarter of 2026.
On payment size, Chainalysis said that over the 16 months ending in September 2026 “over 87% of payments fall between a hundredth of a cent and a dime, with the most common size between one and ten cents.”
How x402 wallets changed, as Chainalysis measured them: the share funded again within 30 days rose from 70.6% to 89.3%, and the share paying within an hour of funding from 4% to 67%.
What Was Posted on X
Aaron Levie of @Box on how AI agents with wallets could create new business models for the internet:
“I’ve always wanted micropayments for things online… I’ve appreciated why it’s a bad businessmodel, but I want it.”
“Now it’s probably a good businessmodel. Let’s say you had paywalled, really proprietary data. You’re Bloomberg. You’re a medical institution. You could imagine there’s maybe a hundred million agents that would be tapping into that data.”
“All of a sudden, a penny a transaction starts to make a lot more sense than it did when you were capped at 20 million [knowledge workers] that maybe were interested in your content.”
“We probably have new business models for tools and APIs… We probably want an agent to be able to have a budget and make any API call it wants.”
“All of a sudden you have this new revenue stream, which is these agents showing up, just transacting on your system.”
a16z posted a clip on 15 October 2026 of Box CEO Aaron Levie arguing that agents with wallets could make micropayments a workable businessmodel: a holder of proprietary data such as Bloomberg or a medical institution might see ‘maybe a hundred million agents’ tapping it, so ‘a penny a transaction starts to make a lot more sense’ than selling to the roughly 20 million knowledge workers he says might want the content. These are his estimates, not measured figures; the x402 transaction counts in this piece come from Chainalysis.
Why the Payments Are So Small
Chainalysis gives the mechanism: transactions worth less than a dollar “don’t make fiscal sense for a credit card network to process”, while stablecoins that cost less than a cent to settle make it “economically viable for agents to spend tiny amounts of money”, such as paying for a data pull through an API call.
Circle, the issuer of the USDC stablecoin, described the split to Chainalysis as “card rails for consumer-prompted agent payments, stablecoin rails for machine-to-machine settlements that are too small for card economics.”
We reported on 13 October that a new Shop skill lets ChatGPT and Claude buy with Shop Pay. As we read it, that is the consumer-prompted side of the split Circle describes.
BlackRock’s head of digital assets research, Will Su, told Chainalysis the firm is watching agentic commerce closely, “with AI agents potentially creating meaningful demand for frequent, low-value payments around the clock.”
Wallets That Keep Coming Back
Chainalysis measured how often wallets were funded again within 30 days of their first funding. That share rose from 70.6% for wallets first funded in Q3 2025 to 89.3% for those first funded in Q2 2026, it said, counting only wallets that sent on two or more separate days.
It also measured how quickly new wallets start paying. In late 2025, wallets were at the median 73 days old when they made their first x402 payment; by Q3 2026 that gap had dropped to 18 minutes, Chainalysis said.
The share of wallets paying within an hour of being funded went from 4% in Q3 2025 to 67% in Q3 2026, according to the report.
What the Agents Are Buying
Chainalysis said early x402 activity was dominated by fads such as memecoins, but that it has since tracked millions of transactions into information services such as code review, financial intelligence, and data and search APIs. “One of the largest categories for x402 transactions is model inference,” it wrote.
US-funded wallets hold a 30% share of the agent-to-services transactions Chainalysis could geolocate. It said American wallets accounted for 88.1% of geolocated model inference transactions, 41.8% of payments to agent frameworks and tool catalogs, and 23.2% of payments to data and search APIs.
Chinese-attributed wallets were most prominent in security: Chainalysis said nine out of ten transactions it tracked into “Security and code analysis” came from Chinese-linked wallets, mostly into AgentLISA, while those wallets took 0.1% of model inference.
The Structural Read
As we read it, the pricingmodel follows the rail: payments of a few cents per call make per-use pricing practical for data, code review and inference services that would otherwise sell seats or monthly plans.
As we read it, the behaviour figures matter more than the transaction count: Chainalysis says 89.3% of the latest wallet cohort was funded again within 30 days and that the median gap to a first payment fell to 18 minutes, which it describes as a recurring budget line rather than a demonstration.
As we read it, the buyer is often another model: Chainalysis says model inference is one of the largest x402 categories and that US wallets account for 88.1% of the geolocated inference transactions.
Circle, as quoted by Chainalysis
“card rails for consumer-prompted agent payments, stablecoin rails for machine-to-machine settlements that are too small for card economics”
Three Implications
API AND DATA SELLERS Chainalysis says it tracked millions of x402 transactions into code review, financial intelligence, and data and search APIs, sold in payments that mostly fall between a hundredth of a cent and a dime.
PAYMENT NETWORKS Circle told Chainalysis that card rails fit consumer-prompted agent payments and stablecoin rails fit machine-to-machine settlements too small for card economics.
COMPLIANCE TEAMS Chainalysis argues that the reviewable checkpoints are the funding event, the counterparty and the behavioural pattern, because individual payments are too fast, small and frequent to screen one by one.
The framework puts it this way: “SaaS is charged for access to tools. AGaaS charges for the execution of outcomes.”
As we read it, x402 is the billing layer that modelneeds: an agent paying a few cents for each data pull or inference call is paying for executed work, call by call, rather than for a seat.
What US-funded AI agents buy on x402, as Chainalysis reported it on 14 October 2026: US wallets made 88.1% of the geolocated model inference transactions and 30% of all geolocated agent-to-service payments.
What Is Not Established
Chainalysis does not give a dollar value for x402 payments in the section we read, so the size of this market in money terms is not established here. Chainalysis also says agentic commerce is still in its “experimentation phase”, as multiple experts it spoke to put it.
The report and Chainalysis’s post date the 138 million figure differently: the report says it was reached by the end of its 2026 reporting period, which ends 1 July 2026, while the post puts 138.5 million in Q3 2026. Fortune’s figure of about 1,500 in mid-2025 is Fortune’s; Chainalysis’s post gives 58.8k for Q3 2025.
Chainalysis said on 14 October 2026 that more than 138 million x402 transactions had occurred on Base, that over 87% of x402 payments were between a hundredth of a cent and a dime, and that newer wallets are funded again more often and start paying within minutes.
A note on sourcing. We read the agentic-commerce section of Chainalysis’s 2026 North America report (published 14 October 2026) in full, along with Chainalysis’s own post on X and Fortune’s report of the same day; the figures and quotations here come from those texts, and the Fortune figure is attributed to Fortune. We did not see Chainalysis’s underlying data or its method for identifying agent wallets.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur, HackerNoon and Search Engine People. Find him on LinkedIn and Substack. See how we source.
Scroll to Top
Discover more from FourWeekMBA
Subscribe now to keep reading and get access to the full archive.
