What Happens When AI Agents Start Paying for APIs?

Model Context Protocol (MCP) is a standard, originally proposed by Anthropic and now adopted by most major AI vendors, that lets an AI model talk to external tools in a uniform way. Instead of every tool inventing its own integration, the tool exposes itself as an MCP server, the model exposes itself as an MCP client, and they negotiate what's available. Today's MCP setup is mostly self-hosted tools, with the user authenticating once and the model calling for free thereafter. Bazaar MCP changes the second half of that sentence.

The current pattern is: you hold accounts with a few API providers, the agent uses your credentials. That works for a handful of tools. It breaks the moment the agent decides on the fly that it needs a tool you don't have. Two examples make the limit concrete.

Discovery-driven tasks. You ask the agent to summarise UK rainfall data for last week. It works out it needs Met Office data, which has a paid tier. You haven't set up an account. The agent stops and waits for you to sort the credentials out. Multiply that across every paid data source and the agent's autonomy disappears.

One-off, low-cost calls. A task needs three geocoding lookups and one satellite tile. Each costs pennies. The cost of setting up an account, putting a card on file, and storing the credentials safely is far higher than the value of the data. Most users will just do it manually rather than absorb that overhead.

The pitch behind agent-side payments is that both of those problems vanish if the agent itself can spend small amounts on the user's behalf, with the user setting a budget and the provider getting paid per call. The technical name for that pattern is micropayments. It has been tried for human-paid web content for two decades and never quite worked; the argument for trying it again is that agents are far better suited to per-call decisions than humans were.

Three pieces, layered on top of vanilla MCP.

A marketplace, not just a protocol. Bazaar surfaces what tools exist, who runs them, what they cost per call, and what permissions they need. The agent can shop without you pre-installing every server it might want.

Payment built in, not bolted on. Each tool can declare a price per call. When the agent uses it, the user's wallet is debited automatically using Coinbase's x402 stablecoin protocol. The user sets the total budget and per-tool caps; everything below those caps is silent.

Auth without accounts. Because the payment is the access, there's no user signup, no API key, no OAuth flow. The agent calls the tool with a payment-attached request and gets a response. The friction that today kills micro-transactions is moved off-path.

The crypto angle (Coinbase, stablecoins, on-chain settlement) is doing real work here, but the surface for the user looks like a normal pre-paid budget: top it up, set the cap, watch the spend. Whether the rails underneath are blockchain or traditional payment infrastructure doesn't really change the user experience.

Glasser (a pay-as-you-go data broker for AI agents) takes the opposite route to Bazaar. Instead of a marketplace of new tools paid in stablecoins, it resells access to data providers you already know: Ahrefs, Semrush, Apollo, Hunter, People Data Labs, Google Maps, Keepa and dozens more. According to its own agent instructions, one key opens more than 2,000 paid endpoints across over 80 providers, and the response is the provider's own output, passed through unchanged.

The billing model is a single pre-paid balance in US dollars, topped up by the human in Glasser's web console. The agent cannot add funds itself. Each call works in three steps:

• Search. The agent looks for an endpoint that fits the task. Searching is free.
• Inspect. The agent reads the endpoint's price and its charge rules before spending anything. Inspecting is also free, and some endpoints charge nothing when a search comes back empty.
• Run. The call goes through, the balance is debited, and the agent reports the charge alongside the provider's raw answer.

It plugs into local coding agents (Claude Code, Codex, Cursor, Gemini CLI) through a command-line tool, and into hosted assistants such as Claude on the web through an MCP server. New workspaces get one US dollar of free credit to try it.

For light, occasional use, clearly yes. Glasser prices everything in US dollars. At the time of writing, its homepage lists a Semrush keyword-metrics lookup at US$0.0011 per call, against US$139 a month for a Semrush plan that doesn't include API access. An Apollo company search is listed at just under two US cents a call, against US$65 per seat per month for Apollo's Basic plan.

Run the sums on Apollo and the subscription only wins past roughly 3,300 company searches a month. A small agency checking a few competitors a week, or a solo developer building an agent that enriches the odd lead, is nowhere near that. Below the break-even point, per-call billing is the cheaper option, and you skip a pile of separate sign-ups, invoices and API keys.

Two caveats. Prices are in dollars, so UK users carry currency conversion on top. And per-call prices vary between endpoints and change over time, so the agent's inspect step, not a price you read on a web page last month, is the number that counts.

Both let an agent buy data one call at a time without you signing up to each provider. The differences sit underneath.

Who you pay. With Bazaar, the agent pays each tool directly in stablecoins. With Glasser, you pay Glasser, and Glasser pays the providers. That is a middleman, with the convenience and the dependency that implies.

What's on offer. Bazaar's early catalogue leans towards developer infrastructure and crypto-native data. Glasser's is built from mainstream commercial sources: SEO metrics, company and contact records, social media data, financial filings, property data.

How you fund it. Bazaar needs a Coinbase account and a stablecoin wallet behind the scenes. Glasser is a conventional pre-paid balance with no crypto involved.

For most people building practical agents today, Glasser is the more immediately useful of the two, simply because the data it sells is data people already pay for.

If you're a hobbyist building personal AI tooling, the honest answer is: not yet. Three reasons.

The first is that the catalogue is small. Bazaar at launch has a few dozen tools, mostly aimed at developer infrastructure and crypto-native data sources. The high-value mainstream APIs (Google Maps, Stripe, Companies House, the Met Office) are not on it. Until that changes, agents using Bazaar can do a small set of things very well and most of nothing else. Glasser goes some way to plugging that gap for SEO, company and social data, which is why it is the more practical of the two today.

The second is that you almost certainly already have accounts for the APIs you actually use. Glasser's own instructions tell agents to use a key you already hold before reaching for its paid catalogue, which is the right default. The pain Bazaar solves (the agent reaching for a tool you don't have) is mostly a pain for autonomous agents, not for someone driving Cursor or Claude Code interactively. Manual signups are a one-time cost and you've already paid most of them.

The third is that it's first-mover in a category. Anthropic, OpenAI, Cloudflare and several payment providers are all building something in this space. Picking Bazaar as the standard today is committing early to a protocol that might or might not be where the category lands. For a hobby project, waiting six months costs nothing.

If you are building agents that need premium data only now and then (a competitor brief, an SEO audit, a batch of company lookups), per-call access is worth trying now. The free starter credit covers a fair amount of experimenting at fractions of a cent per call.

The category does deserve attention, though. The shift from "agent that asks me for credentials" to "agent that has a budget" is the bigger version of the same step desktop computing took when it moved from "machine that runs my programs" to "machine that buys things on my behalf". That step happens once, and 2026 is when it starts happening for agents.

Three real risks worth flagging before anyone wires this up to a personal budget.

Runaway agents. An agent that can spend money without asking is an agent that can spend a lot of money very fast if it gets stuck in a loop. The budget caps Bazaar exposes are a real mitigation, but they only work if the per-tool caps are set conservatively. A buggy agent retrying a £0.10 call a thousand times empties £100 from a budget faster than the user can notice. The default should be set-low-and-raise-deliberately, not set-high-and-trust.

Bad-tool risk. If anyone can publish to the marketplace, the marketplace ends up with low-quality, mislabelled, or actively malicious tools. The agent decides which to call based on the tool's own description. A tool that claims to be "UK Companies House lookup, £0.05 per call" but actually returns junk data while charging you is the textbook failure mode. Curation matters far more than the protocol underneath.

Guardrails that live in the prompt. Glasser's agent instructions say to name the per-call price and wait for your go-ahead before any bulk or repeated paid calls, and to retry a dropped call with the same idempotency key so it isn't charged twice. Those are sensible rules, but most of them are instructions to the agent rather than limits the system enforces. The hard limit is the balance itself: top it up in small amounts and an agent that ignores its instructions can only spend what's there. There's also the middleman to weigh up. Every query you run passes through Glasser on its way to the provider, so read its terms before sending anything sensitive, such as client names or personal data.

None of these makes the category wrong. They make the early implementations risky for anything that matters.