x402-bazaar-survey/POST.md at master · rikocr8orh8/x402-bazaar-survey · GitHub

//blob/show" data-turbo-transient="true" />

Skip to content

Search or jump to...

Search code, repositories, users, issues, pull requests...

-->

Search

Clear

Search syntax tips

Provide feedback

--><br>We read every piece of feedback, and take your input very seriously.

Include my email address so I can be contacted

Cancel

Submit feedback

Saved searches

Use saved searches to filter your results more quickly

-->

Name

Query

To see all available qualifiers, see our documentation.

Cancel

Create saved search

Sign in

//blob/show;ref_cta:Sign up;ref_loc:header logged out"}"<br>Sign up

Appearance settings

Resetting focus

You signed in with another tab or window. Reload to refresh your session.<br>You signed out in another tab or window. Reload to refresh your session.<br>You switched accounts on another tab or window. Reload to refresh your session.

Dismiss alert

{{ message }}

rikocr8orh8

x402-bazaar-survey

Public

Notifications<br>You must be signed in to change notification settings

Fork

Star

FilesExpand file tree

master

/POST.md

Copy path

Blame<br>More file actions

Blame<br>More file actions

Latest commit

History<br>History<br>History

148 lines (111 loc) · 8.54 KB

master

/POST.md

Copy path

Top

File metadata and controls<br>Preview

Code

Blame

148 lines (111 loc) · 8.54 KB

Raw<br>Copy raw file<br>Download raw file

OutlineEdit and raw actions

I pulled all 14,865 listings in Coinbase's x402 Bazaar. 3.5% of them have a real user.

Data as of 2026-07-28. The registry is live and will look different by the time you read this —<br>that's the point: re-run the method yourself and see what's changed.

The x402 Bazaar is Coinbase's public discovery registry for machine-payable<br>HTTP endpoints — the thing an AI agent queries to find an API it can pay for per-call instead of<br>signing up for. It's a young, fast-growing catalog: 14,865 listings across 1,561 distinct hosts as<br>of this pull.

I wanted to know: how many of those listings are actually being used, versus just... listed?

The headline number

520 listings (3.5%), across 198 sellers, show organic repeat demand. Everything else — 96.5%<br>of the registry — has either zero usage or usage that doesn't look like a real, repeat customer.

Whole-market organic volume across the entire Bazaar: ~260,000 calls in the last 30 days. At a<br>median listed price of $0.014/call, that's roughly $3,000–6,000/month in real, repeat-usage GMV<br>for the entire registry. Not per top seller — for everything.

And it's concentrated: the top 10 sellers take 89% of that organic volume.

This isn't a "x402 is dead" post. 260K real calls a month from real agents paying real money is not<br>nothing, and Coinbase is putting serious weight behind the protocol. The actual finding is narrower<br>and more useful: the market is real, small, and currently ranked by a metric that doesn't measure<br>demand.

Method — and how to check it yourself

Every listing in the Bazaar carries a public quality field: 30-day call count and 30-day unique<br>payer count. It's returned by the same public, unauthenticated endpoint anyone can hit:

GET https://api.cdp.coinbase.com/platform/v2/x402/discovery/resources?limit=100&offset=0

Nobody appears to be using this field for discovery today — the Bazaar's own UI and the main<br>third-party explorer both rank by raw settled volume, which (see below) is exactly the number that's<br>easiest to inflate without any real usage behind it.

I define organic demand as: at least 3 unique payers, and at least 3 calls per payer on average.<br>That's a low bar — it just asks for some repeat usage from more than one wallet — and only 3.5%<br>of the registry clears it.

I define a one-call-per-wallet pattern as: 50+ unique payers, averaging under 1.5 calls per<br>payer. That's the shape you'd expect from a one-shot registration flow, an airdrop/incentive<br>campaign, or (less charitably) incentive farming — not from an agent that keeps coming back.

Both scripts that produced every number in this post are in the same repo:

pull-bazaar.py — paginates the public registry into a slim JSON snapshot

score-bazaar.py — applies the two definitions above and prints the full breakdown

python3 pull-bazaar.py && python3 score-bazaar.py

Run it today and you'll get numbers close to but not identical to this post's — the registry moves.<br>Run it and get numbers wildly different, and one of us has a bug; open an issue.

What agents actually pay for

The organic-demand list is short and top-heavy. The four largest organic sellers by 30-day calls:

Endpoint<br>Organic calls (30d)<br>Unique payers<br>What it does

x402.twit.sh<br>117,323<br>45<br>X/Twitter search

x402.tavily.com<br>60,560<br>417<br>Tavily web search

stableenrich.dev (Exa)<br>10,539<br>265<br>Exa neural web search

api.exa.ai<br>4,294<br>98<br>Exa search (direct)

These four endpoints alone account for over 190,000 of the ~260,000 organic calls in the entire<br>registry. The pattern across the whole organic list is the same:...