How to price an API for AI agent buyers

Live pricing on our own endpoints, and what settled and what didn't.

Pricing for a machine buyer isn't a discount version of pricing for a human. There's no sales call, no annual-plan upsell, no "talk to us for enterprise pricing." There's a number in a 402 response, and the agent's task logic either pays it or moves on. Here's what we measured running real prices on real endpoints.

Our price ladder, live

We priced the raw, single-purpose data calls at fractions of a cent to a nickel — grid-sizing parameters at $0.05, regime classification at $0.03, a price/momentum check at $0.005. We also built two bundles that combine four calls' worth of signal into one request, priced at $1.00 and $2.50, on the theory that convenience justifies a markup.

What actually settled

Eighteen calls have settled, all on-chain, on Base. Every single one hit the sub-$0.05 raw-data endpoints — grid parameters, market regime, live P&L telemetry. Zero settled calls have hit either bundle, at $1.00 or $2.50, in over a week of both being live and discoverable in the same catalog as everything else. Not "low conversion" — zero.

That's a real, reproducible finding from our own catalog traffic, not a theory: at this stage of the x402 market, buyers (or the agent logic deciding for them) will pay a few cents for a specific fact instantly, and will not pay a dollar-plus for a bundle of the same facts, even when the bundle is objectively the better deal per call. Convenience pricing assumes a buyer who values their own time enough to pay a premium for it. An agent optimizing a task budget doesn't work that way — it wants the cheapest unit of the specific fact it needs, right now.

Why machine pricing behaves differently from human pricing

A human evaluating a $1 API call weighs it against the time it would take to gather the same information manually — that's the whole logic behind "convenience" pricing, and it works on humans constantly. An agent executing a task doesn't have that calculus. It has a budget, a goal, and a decision function that's usually optimizing for cost-per-unit-of-useful-signal, not cost-per-minute-saved. When the same fact is available at $0.03 from one source and bundled at $1 with three other facts from another, the cheaper atomic unit wins even if the bundle is a better deal on paper — because nothing in the agent's loop is pricing in its own "time," the way a human would price in theirs.

That's also why per-call microprices aren't a race to the bottom the way discounting is for human SaaS. There's no customer-acquisition cost to recover with a higher price point, no sales team to fund, no annual contract to protect. The economics of an agent-facing endpoint look more like a utility meter than a subscription: usage-based, small, and only valuable in aggregate if buyers come back.

What we measured on discovery vs. settlement

Our catalog and MCP listings get checked constantly — discovery crawlers, registry indexers, and agent frameworks probing what's available run in the dozens-per-day range on our own logs, most of it unauthenticated handshake traffic rather than paid calls. That's normal and worth knowing before you panic about "low conversion": most of what hits a freshly listed endpoint is infrastructure noise, not a buying decision. The only number we treat as signal is settled, on-chain revenue, which is why every figure on this page traces back to receipts we can point to, not request logs.

Rules we'd give another builder

• Start at fractions of a cent, not dollars. The x402 economy we measured across 25,000+ registered endpoints has a median asking price around $0.02/call. That's not a race to the bottom — it's what a single unit of machine-consumable data is actually worth when there's no human decision cost attached to the purchase.
• Price the atomic fact, not the convenience bundle. If you have five data points, price access to each one separately before you build a "premium" combined tier. Our own data says the combined tier is the thing that doesn't sell yet.
• Don't confuse a low price with low revenue potential. A wallet that returns weekly at $0.03–$0.05 a call is worth more over a quarter than one $1 purchase that never repeats. Optimize for repeat settlement, not per-call size.
• Publish your real numbers. We put $0.235 lifetime revenue on our own homepage-adjacent page (/open). It's a small number. It's also the only number in this entire space that isn't a projection.

Source for our own figures: live receipts.jsonl and revenue_events.jsonl on our production x402 API, reconciled 2026-07-19. Median-price figure: our own 2,000-resource sample of Coinbase's CDP x402 discovery catalog (see /state-of-x402 for method).