← Home · Blog · Ecosystem

Trust filter · live snapshot

x402 Bazaar Market Map

Live per-agent cut, beyond this snapshot: x402 · https://api.onchainagentintel.io/v1/intel/agent/{agent_id} · buy access.

Data as of 2026-09-25T09:01:49Z UTC · refreshed daily by an automated snapshot pipeline (task 0267).

Our snapshot of the public Coinbase CDP x402 Bazaar discovery catalog — cut by CDP’s unique-payer signal, not by raw listing count. 17,438 endpoints indexed from https://api.cdp.coinbase.com/platform/v2/x402/discovery/resources. Snapshot fetched 2026-09-25T09:01:49Z (status: ok; reported catalog total: 17,438).

17,405
≥1 unique payer
99.8% of catalog
926
≥5 unique payers
5.31% of catalog
355
≥10 unique payers
2.04% of catalog
41
≥100 unique payers
0.235% of catalog

Headline: only 926 of 17,438 indexed endpoints (5.31%) report ≥5 unique payers in CDP’s own quality signal over the last 30 days. That is the strong-filter cohort — the rest is long tail.

Roughly 99.8% of the catalog reports at least one unique payer in the last 30 days, so a > 0 test is not itself a trust signal. Payer thresholds tighten the cohort in a way a binary present-or-absent cut cannot. Signal source is CDP’s quality.l30DaysUniquePayers, applied verbatim — nothing self-reported, nothing scored by us.

Concentration headline — both sides of the market, one block. Supply (our live CDP snapshot, 2026-09-25T09:01:49Z): 926 of 17,438 indexed endpoints (5.31%) cross the ≥5-unique-payer threshold in the last 30 days. Demand (independent operator x402-list.com, frozen citation 2026-08-15): top ten services capture 97.9% of settled USDC volume across 423 measured services ($206,753 USDC settled 30d, 5,745 unique buyers). Two operators, two methodologies, same shape: listing is not demand.

Sample-vs-population, stated plainly. The supply-side cohort is a whole-catalog cut of the CDP Bazaar (~15k endpoints, denominator above). The demand-side number is a sample, not the population: x402-list.com's own coverage_of_measured_flow_pct field self-reports as 17.5%, so “top-10 = 97.9% of settled volume” is 97.9% of the volume across their 423 measured services, not of every x402 dollar in the wild. Their top-1 service's history is imported, not natively observed (top_1_service_imported: true) — the 83.3% top-1 share is a directional upper bound on that one service, not a raw handshake count. Full reconciliation: the x402 concentration reconciliation.

Call-weighted view of the same live CDP snapshot. The same ≥5-unique-payer cohort that is 926 of 17,415 indexed endpoints with a payer AND a call count (5.32% endpoint-weighted) carries 600,882 of 749,440 total 30d requests indexed (80.2% call-weighted). The two shares live on different denominators — endpoints with both fields present vs total 30d requests — and are stated together on purpose: endpoint share and call share answer different questions and one without the other misreads the shape. Endpoints with exactly one unique payer — 12,813 of 17,415 (73.6% endpoint-weighted) — carry 88,721 of 749,440 30d requests (11.8% call-weighted); the median 30d call count for that cohort is 1 — the typical 1-payer endpoint has been touched almost trivially.

Wording contract, sitting under the CDP conditional caveat above. l30DaysTotalCalls is a request count, not settled USDC value — nothing here is “volume” in a dollar sense and no conversion is applied. l30DaysUniquePayers counts distinct paying wallet addresses; a single address may be an operator wallet, a funding wallet, or a canary, so a 1-payer endpoint is indistinguishable from self-dealing from on-chain data alone — not proof of it, and not proof of a real external customer either. The CDP Bazaar catalog is one x402 discovery source, Base-weighted; both denominators above are conditional on being indexed by CDP.

Long-form treatment of the endpoint-vs-call inversion, published on the same snapshot cadence: Report 10 — The Call-Weighted x402 Demand Cut.

Unique-payer distribution (last 30 days)

Full histogram, exclusive buckets — counts sum to the indexed total. The long tail on the right is the trust cohort; the mass on the left is what makes a bare “> 0 payer” filter meaningless.

Payer bucket (30d) Endpoints Share
no CDP quality data230.132%
0 payers (30d)100.057%
1 payer (30d)12,81373.5%
2 payers (30d)2,38413.7%
3–4 payers (30d)1,2827.35%
5–9 payers (30d)5713.27%
10–49 payers (30d)2871.65%
50–99 payers (30d)270.155%
100+ payers (30d)410.235%

Reputation footprint of the earning cohort (Base, snapshot 2026-09-01)

Endpoints that earn money on Base and endpoints that carry an on-chain reputation footprint are almost entirely disjoint. Of the 14,338 Base x402 endpoints in this snapshot that took ≥1 paid call in the last 30 days, 1,291 (9.00%) resolve to a registered ERC-8004 agent owner by exact pay_to → owner match, 9 (0.063% of the earning set; 0.697% of the resolvable subset) carry any on-chain reputation feedback, and 0 carry commerce-backed reputation — feedback backed by a real ERC-8183 job outcome rather than a bare thumbs-up. The same shape at the publisher level: 26 distinct pay_tos, 4 with any feedback (one event each, one unique client apiece), 0 commerce-backed. Every share here is a lower bound (pay_to may be a treasury wallet distinct from the registration owner; we index the ERC-8004 registry on Base and Ethereum only; per-publisher dedup capped at 50 rows so a single bulk-mint publisher cannot flood the resolvable set — disclosing that cap is what makes the resolvable count auditable). Full method and predicate: /how-do-you-know-which-agent-is-safe-to-pay. Dedicated dataset with the publisher-level cut, JSON + CSV: /x402-paid-but-unrated. That gap is where our paid intel lives.

Since previous snapshot

15
Added
0
Removed
0
Price-changed

Since previous snapshot at 2026-09-25T08:30:49Z — this delta updates on every daily refresh of the market map.

By network

Which chains the Bazaar catalog settles on. An endpoint can list multiple networks, so the sum can exceed the endpoint count.

Network CAIP-2 Endpoints
Baseeip155:845316,988
Solanasolana:5eykt4UsFv8P8NJdTREpY1vzqKqZKvdp5,487
Polygoneip155:1373,048
Arbitrumeip155:421612,447
eip155:84532eip155:845321,096
eip155:143eip155:143857
eip155:4663eip155:4663678
algorand:wGHE2Pwdvd7S12BL5FaOP20EGYesN73ktiC1qzkkit8=algorand:wGHE2Pwdvd7S12BL5FaOP20EGYesN73ktiC1qzkkit8=664
xrpl:0xrpl:0649
Optimismeip155:10630
Avalancheeip155:43114613
eip155:1329eip155:1329610
eip155:42220eip155:42220604
Stellarstellar:pubnet602
basebase350
eip155:480eip155:480176
eip155:196eip155:196134
BNB Chaineip155:5696
algorand:wGHE2Pwdvd7S12BL5FaOP20EGYesN73kalgorand:wGHE2Pwdvd7S12BL5FaOP20EGYesN73k79
solanasolana69
eip155:999eip155:99963
cosmos:noble-1cosmos:noble-154
Solanasolana:EtWTRABZaYq6iMfeYKouRu166VU2xqa139
Ethereumeip155:131
eip155:4326eip155:432623
nano:mainnetnano:mainnet15
eip155:5042eip155:504214
eip155:42793eip155:4279310
eip155:1187947933eip155:11879479338
eip155:130eip155:1308
eip155:146eip155:1468
zeroclick:cardzeroclick:card7
eip155:1440000eip155:14400006
hyperliquid:mainnethyperliquid:mainnet6
animica:1animica:15
base-sepoliabase-sepolia5
stacks:1stacks:15
xrpl:mainnetxrpl:mainnet5
hedera:testnethedera:testnet4
eip155:50eip155:503
cosmos:gonka-mainnetcosmos:gonka-mainnet2
eip155:5042002eip155:50420022
aws:baseaws:base1
eip155:11142220eip155:111422201
eip155:80002eip155:800021
lightning:mainnetlightning:mainnet1

By category

Lightweight keyword bucketing from each endpoint's description, service name, URL, and tags. "uncategorized" is anything the keyword rules didn't match.

Category Endpoints
uncategorized7,507
payments5,743
web-scrape1,774
data-analytics857
ai-inference530
compliance-risk507
chain-rpc461
reputation-intel59

Price distribution (USDC)

Cheapest USDC-denominated accepts[] entry per endpoint. Non-USDC legs (SPL, Stellar assets, non-stablecoins) don't show up here. Full percentile ladder, histogram, and per-bucket CSV live at /what-does-an-x402-api-call-cost — the canonical price-census answer page.

17,142
USDC-priced endpoints
$0.010
Median price
$0.0000
Cheapest
$1,000.00
Most expensive

Most expensive endpoints

# Host Service Min USDC price
1api.bitrefill.comBitrefill$1,000.00
2api.jamton.networkHerculis Gold Coin$1,000.00
3mudko.comAgent Ready Kit$450.00
4mudko.comBazaar Keepalive (prepaid year)$190.00
5laso.financeLaso Finance$103.80
6mudko.comBazaar Listing Package$99.00
7orcpin.devOrcpin$50.00
8orcpin.devOrcpin$50.00
9www.productclank.comProductClank Content Campaign$50.00
10agentpaystore.comActivate a 30-day blimp ad slot on all AIARENA pages. $40 USDC via x402.$40.00

Top hosts by endpoint count

Concentration — a small number of hosts publish a large share of the catalog.

# Host Endpoints
1agentshelf.syntexa.ch1,182
2intel.rallylive.ca840
3agent402.tools602
4pay.edge-agents.ai582
5saylorinnovations.com376
6twin.unykorn.org352
7api.x402node.dev340
8x402-factory.com205
9402.com.tr161
10api.strale.io160
11minia2a.uk153
12api.agentstools.dev146
13agents.x402stock.xyz138
14api.delx.ai125
15atlantikland.halowerk.com125
16baltikum.halowerk.com125
17mitteleuropa.halowerk.com125
18nordland.halowerk.com125
19ostland.halowerk.com125
20suedland.halowerk.com125

Methodology & source

Source: https://api.cdp.coinbase.com/platform/v2/x402/discovery/resources — the public Coinbase CDP x402 Bazaar discovery API. No API key required. On-Chain Agent Intel fetches the full catalog once per day, normalizes each item into a stable record shape (network, scheme, payTo, USDC price, quality signal), and diffs against the previous snapshot to produce the delta panel above. The trust-filter cohort uses only fields CDP itself publishes on the endpoint — nothing self-reported, nothing scored by us.

Why unique payers, not total calls. CDP publishes both quality.l30DaysUniquePayers and quality.l30DaysTotalCalls on every indexed endpoint, and both are non-zero on the same near-universal share of the catalog. This page reports the unique-payer signal because it’s harder to inflate — a single client hammering an endpoint bumps calls without changing the payer count, so payers is the tighter filter for real usage breadth. Both signals come from CDP; neither is de-noised by us.

Snapshot cadence is daily — every fresh build of this page reads the latest fetch. This is a live-updating dataset page, not a one-off report. For the long-form treatment of what this data means for the ERC-8004 / agent-intel corner of the Bazaar, read Report 06 — The x402 Agent-Intelligence Market Map. For the whole-catalog demand-concentration headline as a standalone citable finding, see Report 07 — Only 5.31% of x402 endpoints have ≥5 unique payers in 30 days.

On-Chain Agent Intel is not affiliated with Coinbase. The Bazaar catalog is published by CDP under the URL above; we redistribute a normalized cut of it as a live-updating reference page.

Cite this dataset

Headline finding. As of 2026-09-25T09:01:49Z, 926 of 17,438 (5.31%) x402 endpoints in the public CDP Bazaar catalog reported ≥5 unique payers in the last 30 days.

Citation. On-Chain Agent Intel. "x402 Bazaar Market Map — daily trust-filter snapshot." Snapshot 2026-09-25T09:01:49Z. https://onchainagentintel.io/x402-bazaar-market-map

BibTeX. Copy-paste for LaTeX / biblatex / Zotero:

@misc{ocai_x402_bazaar_market_map,
  author       = {{On-Chain Agent Intel}},
  title        = {x402 Bazaar Market Map — daily trust-filter snapshot},
  year         = {2026},
  month        = {09},
  howpublished = {\url{https://onchainagentintel.io/x402-bazaar-market-map}},
  note         = {Snapshot: 2026-09-25T09:01:49Z. License: CC BY 4.0.},
}

License. Published under CC BY 4.0 — free to reuse, including in commercial work, provided you attribute “On-Chain Agent Intel” and link back to https://onchainagentintel.io/x402-bazaar-market-map.

Machine-readable exports. JSON · CSV — the same numbers rendered above, refreshed daily alongside this page.

Also useful: Ethereum chain brief · Base chain brief · Commerce-backed agents · API docs · llms.txt