Stats: paid agent tools
Live numbers from the current Agentropolis index, with wash trading filtered out. Frozen weekly editions: State of paid agent tools (latest: 2026-40) · RSS · JSON: /api/v1/stats · per-category pages: /categories
Data snapshot: Oct 2, 2026, 12:17 PM ET · window: rolling 30 days ending at data_as_of · ISO week 2026-40
Headline (wash-filtered)| x402 services with verified paid usage | 466 |
| Settlements, last 30 days | 370,307 |
| Measured volume, last 30 days | $44,378.7 USDC |
| Verified buyers, summed per service* | 15,328 |
| Median verified buyers per service | 9 |
| Median lowest listed price | $0.01 per call |
| Listings excluded for wash signals | 112 of 842 x402 services tracked |
| Top 10 services' share of summed buyers | 37.42% |
* A wallet that pays several services is counted once per service, so this is not the number of unique buyers.

By category
| Category | Services with paid usage | Verified buyers* | Settlements 30d | Volume 30d | Median price | Excluded (wash) |
|---|
| Crypto & on-chain data | 70 | 5,574 | 82,298 | $31,416.7 | $0.0075 | 14 |
| Security & verification | 101 | 2,498 | 19,271 | $203.81 | $0.01 | 20 |
| Media: speech, video & image | 44 | 1,722 | 18,901 | $3,335.52 | $0.005 | 12 |
| Scraping & extraction | 32 | 968 | 5,266 | $651.16 | $0.01 | 9 |
| LLM inference | 43 | 936 | 115,020 | $5,186.88 | $0.01 | 11 |
| People & company enrichment | 38 | 917 | 67,143 | $1,824.12 | $0.01 | 23 |
| Social data | 24 | 856 | 24,404 | $472.37 | $0.01 | 5 |
| Web search | 36 | 626 | 13,682 | $189.22 | $0.005 | 5 |
| Finance & markets | 23 | 502 | 7,020 | $15.26 | $0.03 | 2 |
| Other | 23 | 272 | 10,958 | $160.96 | $0.01 | 4 |
| Compute & infra tools | 9 | 138 | 3,094 | $80.27 | $0.1 | 1 |
| Off-ramp, gift cards & payouts | 4 | 134 | 1,037 | $830.25 | $0.06 | 0 |
| Travel | 7 | 68 | 791 | $7.77 | $0.05 | 3 |
| Reports & research | 7 | 54 | 943 | $3.83 | $0.01 | 2 |
| Developer tools | 1 | 32 | 370 | $0.09 | $0.02 | 0 |
| Weather & geo | 1 | 16 | 84 | $0.09 | $0.001 | 1 |
| Productivity & docs | 2 | 12 | 20 | $0.15 | $0.002 | 0 |
| Commerce & purchases | 1 | 3 | 5 | $0.25 | $0.05 | 0 |
Top 10 services by verified buyers
Price bands (lowest listed price, services with paid usage)
| under $0.001 | 3 services |
| $0.001 to under $0.01 | 189 services |
| $0.01 to under $0.10 | 205 services |
| $0.10 and up | 69 services |
Index coverage
38,952 listings: MCP servers 37,979, paid HTTP APIs 970, kits 3. Sources: Official MCP Registry 37,979, x402-list 842, MPP registry 137, first-party 3. Discovery files found: x402 manifest 576, OpenAPI 705, llms.txt 669, A2A card 284.
Settlements Agentropolis verified on-chain
3 listings · 0 independent settlements in 30 days · 3 disclosed test purchases (not counted). Our own three kits have only disclosed test purchases so far: no independent sales yet.
Methodology
- Scope: x402 services listed on x402-list.com with on-chain measured USDC settlements in the 30 days before the data snapshot. MPP services and MCP servers are counted in the index totals but have no settlement data yet.
- Wash filter (hub-trust-v0.1): a service is excluded when one buyer makes up 80% or more of volume (10+ settlements), when 2 or fewer buyers account for $50+, or when there are 1.5 or fewer calls per buyer across 50+ buyers (wallet-farm pattern). Excluded services count toward 'excluded for wash signals' only.
- Verified buyers are summed per service: a wallet that pays five services is counted five times. The sum is therefore not the number of unique buyers in the market.
- Prices are the lowest listed price per call or unit from public metadata; the seller's live HTTP 402 challenge is authoritative and actual paid amounts can differ.
- Windows are rolling 30 days ending at the data snapshot time, not the calendar week in the report name. The week is the ISO week of the snapshot.
- Agentropolis-verified settlements are payments we confirmed ourselves by reading Base receipts to a payTo the seller published. Purchases by the seller, the operator or Agentropolis (test sales) are shown but never counted.
- Volume is the measured USDC value of settlements. It can differ from listed price x settlements (tiered prices, discounts, test-priced calls).
- Coinbase x402 Bazaar / Agentic.Market data is not used or republished (their terms forbid aggregation).
Caveats
- Wash flags are signals, not accusations; sybil farms using many funded wallets can still pass.
- x402-list coverage is incomplete, so totals are a lower bound for x402 activity.
- Services that share one payout wallet (marked shared payout) can show identical numbers, because settlements are measured per wallet; each listing is still counted separately.
- Not financial advice.
Attribution. Paid-usage data: x402-list.com (CC BY 4.0, https://creativecommons.org/licenses/by/4.0/), on-chain measured USDC settlements. Wash filtering, aggregation and categories: Agentropolis (hub-trust-v0.1).
Cite as: Agentropolis, "State of paid agent tools, 2026-40", https://agentropolis.io/reports/2026-40. Paid-usage data: x402-list.com (CC BY 4.0, https://creativecommons.org/licenses/by/4.0/), on-chain measured USDC settlements. Wash filtering, aggregation and categories: Agentropolis (hub-trust-v0.1).
Embed the chart: <a href="https://agentropolis.io/stats"><img src="https://agentropolis.io/stats/chart.svg" alt="x402 paid usage by category (Agentropolis)"></a>
Preview. Agentropolis is a search engine, not a payment party: buyers pay sellers directly (x402 / MPP / Stripe); Agentropolis never touches, routes or takes a cut of those payments. Data: Official MCP Registry (CC0), tempoxyz/mpp (MIT/Apache-2.0), x402-list.com (CC BY 4.0), sellers' own discovery files. Seeded listings are unclaimed.
Sources & terms ·
Roadmap ·
Privacy ·
llms.txt ·
A2A card ·
Reports RSS · Contact:
hello@agentropolis.io