Author: jessy
Translated by: Jiahuan, ChainCatcher
For the past year, I have been dedicated to building infrastructure for the Agent economy, having exchanged ideas with teams from Stripe, Visa, Coinbase, Google, and dozens of startups driving Agent commerce. I've combed through the entire industry, launched products, and searched for product-market fit.
There is currently no real demand, and startups face numerous structural issues when entering this space.
Last month, Stripe launched 288 new products at the Sessions conference, and their Agent documentation garnered nearly 40% of the total documentation views. Their Agent commerce marketplace has over 1,000 activated merchants. However, at the Sessions conference, the number of registered Agents making transactions was only in the single digits.
Visa mentioned that their Agent payment tokens (tokenized payment credentials bound to an Agent for making payments on behalf of a user) currently require 3 to 9 months for KYC approval and essentially demand a minimum revenue threshold of $250 million to qualify. Today, only companies at the scale of Amazon or Walmart can complete this identity verification loop.
Coinbase reported that as of April, there were 69,000 active Agents and 165 million transactions on the x402 protocol. But independent on-chain analysis shows the actual daily transaction volume is around $17,000, with about half being test transactions (according to CoinDesk, March 2026).
Agent to Merchant
We built shop.fast.xyz to directly validate the real-world application of drop-shipping style commerce. It includes real products, merchants, and transactions.
For most product categories, the current AI shopping user experience is completely inferior to traditional e-commerce. When you buy clothes, electronics, or furniture, you want to see pictures, browse various options, and compare them side-by-side.
The conversational format of a chatbot is actually a step backward. You are essentially replacing a rich visual interface with pure text dialogue, and humans are fundamentally visual shoppers.
Agents excel in areas where we thought they would struggle. They can understand user needs and handle instructions like "something like this but cheaper" well. The model layer works.
But it cannot replace the experience of browsing ten products side by side and picking one. The chat interface can be enhanced with carousels and interactive displays, but at that point, you're just rebuilding an e-commerce frontend within a chat window. For visual-driven comparison shopping, we haven't found a compelling reason to believe a chat interface is better than a native e-commerce interface.
We see real demand from merchants, but it's a defensive demand.
Merchants want their stores to be queryable by Agents. Not because current customers are buying through Agents, but because they fear being left behind if this becomes the mainstream channel.
It's an "Agent Engine Optimization (AEO)" strategy, but for now, it's just a nice-to-have, not a necessity. Merchants are preparing for a wave that hasn't arrived.
Conversational commerce can indeed improve the experience in certain scenarios: high-frequency, low-decision-cost purchases where users already know exactly what they want. Food delivery is the most obvious example. The market is huge, frequency is high, decisions are quick ("order me Pad Thai from my usual place"). Conversational Agents have a chance here.
But large food delivery platforms haven't opened their APIs. The only path is "computer use": having AI navigate the app visually like a human. This is slow, fragile, and the inference costs for a $15 lunch order simply don't add up.
Another opening lies in: certain stores with extremely convoluted UI navigation that is a pain to use. Layers of discounts, promo codes, loyalty programs, and confusing checkout flows.
An Agent that understands "use my coupons, deduct my reward points, find the cheapest shipping, operate in my native language" can simplify the currently terrible experience. This is particularly important for elderly users, non-native speakers shopping on foreign sites, or in very specific scenarios with niche needs.
Both of these openings require massive consumer-facing (B2C) distribution channels. You're competing with DoorDash (the largest US food delivery platform, holding 56% market share) and Amazon for user entry points.
Consumer-scale distribution favors the giants. The supply side of drop-shipping commerce is ready, while the demand side is constrained by user experience and distribution channels. Building more infrastructure doesn't solve these two issues.
Agent to API
We spoke with dozens of developers about their actual payment needs. The situation is almost eerily consistent: Agent-to-API usage today is constant, covering compute, inference, and data sources. Developers already have subscriptions, stored API keys, and billing relationships with core vendors.
The typical stablecoin argument is: on Stripe, the minimum effective cost for credit card processing is around 2.9% plus 30 cents, making API calls under one dollar uneconomical. But for current low transaction volumes, prepaid balances solve this. Developers top up their accounts in advance, and the problem is gone.
The deeper issue is the supplier market. Most mainstream SaaS companies don't want to offer fractional-penny, ephemeral API access. Their business model is multi-year enterprise contracts. Companies whose revenue relies on large commitment contracts will resist pricing mechanisms that circumvent their existing model.
Machine commerce is structurally a long-tail market, including smaller services, niche data sources, individual developers, and MCP servers. Protocols like MPP and x402 are perfectly suited for this niche.
But by definition, this is a market serving advanced users with special needs, and historically, developers are one of the least willing to pay groups.
When Stripe Projects launched, it partnered with 32 vendor partners like Vercel, Supabase, Cloudflare, Twilio, etc., covering most tools developers use to build and deploy software, all accessible via existing billing systems. The top-of-stack needs of developers are already met.
The opportunity for new payment rails exists for everything beyond these top 30 services: The opportunity is real, but its scale is inherently far smaller than the impressive numbers suggest.
The same pattern applies to content acquisition. Agents are already constantly scraping and summarizing articles, and publishers are fighting back.
But when content monetization arrives at scale, it will happen through CDN providers already sitting between publishers and the internet (Cloudflare has launched AI auditing tools for this), or through large-scale licensing agreements between publishers and AI labs.
This infrastructure opportunity will ultimately flow to the giants who already own the distribution channels.
Agent to Agent
The Agent-to-Agent business model is a long-term vision, currently almost entirely theoretical, with no one achieving meaningful transaction volume. Startups are tackling the core problems: Agent discovery, trust establishment, term negotiation, and dispute resolution.
When this transaction structure truly materializes, it will be fundamentally different from existing payment rails. Neither side of the transaction involves a human identity. Latency is sub-second. Funds from fractions of a cent to millions of dollars flow in the same pipeline.
Furthermore, there are multi-party settlement mechanisms that don't fit the bilateral buyer-seller model assumed by existing payment rails. When this happens, we believe it will come quickly and at large scale.
This is a long-term bet on dedicated settlement infrastructure, and it's real. But a "real long-term bet" and a "current market" are two different things.
For months we were also among those evangelizing this market and built comprehensive infrastructure around it over the past few years. With our distributed network, we could theoretically scale to over 1 billion TPS with sub-50ms latency and 10ms average finality. But we have to meet the market where it actually is right now.
Agent to Finance
This is arguably the only category with existing demand. The customer base already exists and is willing to pay. Today, fund managers, finance teams, and DeFi users pay for financial tools. Embedding AI into existing workflows is a natural product evolution.
Agent finance also creates entirely new behaviors. Agents that can autonomously monitor and rebalance hundreds of positions in real-time operate in ways humans cannot replicate manually. This isn't just automation; it's a qualitative capability upgrade.
The challenge is the competitive landscape. The finance industry is heavily regulated and relies heavily on existing business relationships. Incumbents have licenses, compliance infrastructure, and client relationships. Startups can find niches in less regulated areas (like DeFi), areas where incumbents move slowly, or where AI creates capabilities incumbents don't have.
But compared to the other three categories, the competitive dynamics here are more favorable to established companies, because layering AI on top of existing products and customer bases is far easier than the reverse.
The Real Battle
So why is everyone still building this stuff? There are two reasons.
First is motivation. Industry giants have ample cash flow to bet on a future that may take years to materialize. For them, the cost of entering five years early is a rounding error, while the cost of entering one year late is existential. So they have to build.
Second is cognitive bias. When your core business is payments, every problem looks like a payment problem. The Agent economy needs a payment layer, so build that payment layer.
But payments are just one piece of a much larger puzzle. The real challenge isn't moving money between Agents; it's coordinating work between Agents and humans, verifying that work, and settling the results. Payment is just part of settlement. Settlement is just part of coordination. And coordination is the real prize.
Coordination at scale naturally begets settlement mechanisms as a hard requirement. Payment is just one instrument in that symphony, not the entire composition. The companies that solve coordination will swallow payments, not the other way around.
Most incumbents are building defensively for a future where machines transact at massive scale. Because their funding runway is infinite, timelines aren't critical for them.
But startups don't have that luxury. We have to go find where the market actually is; we can't just wait for the wave to hit.
A year of building has led us in an unexpected direction. There, market activity is real, growing rapidly, and underserved. It falls outside these four categories we outlined.
