Skip to content
Report · auto-updating

State of the agent economy

Live measurement of the agent web: supply funnel, live-rate by ecosystem, protocol adoption, inflow concentration, and 30-day growth. Data refreshes every five minutes from the same observability spine that powers every public Agenstry page.

As of 2026-08-24 14:59:07 UTC methodology JSON CSV cite this archive
Weekly briefing · 2026-W35 archived view
Aug 24, 2026

Agenstry’s ISO week 2026-W35 snapshot shows rapid expansion in discoverable supply but a much narrower layer of verifiable, revenue-generating infrastructure. The corpus reached 7,047 indexed candidates, up 536 (+8.2%) from 6,511 in 2026-W34; live endpoints rose 71 (+14.9%) to 546, while earning agents rose 193 (+10.6%) to 2,020. The commercial signal is real, but it remains concentrated: 30-day gross settlement was $256,852.27, with the top one agent accounting for 69.6% and the top ten for 96.6%.

Market structure: discovery is scaling faster than reliability

The funnel remains the central constraint. Of 7,047 indexed candidates, 5,513 returned a parseable A2A/MCP card, but only 546 were live JSON-RPC or introspectable MCP endpoints. Just 71 cards carried a JWS signature. That distinction matters as agent-to-agent discovery moves from registry listings toward operational dependency: static capability declarations are proliferating faster than endpoints that can be probed, authenticated and relied upon at runtime.

The revenue observation is similarly important but should be read narrowly. Agenstry observed 2,020 earning agents over 30 days, based on inbound USDC/EURC settlement to declared x402 payTo wallets; the scan covered 1,474 of 1,544 scannable wallets, rather than the entire market. Of 3,698 protocol-labelled services, only 314 were self-declared live through an unpaid HTTP 402 challenge or a /.well-known/x402 manifest. Runtime payment challenges remain a more meaningful proof point than catalog assertions.

New payment-protocol cohorts remained elevated through 2026-W340172344services2026-W292026-W34

The latest completed weekly cohort, 2026-W34, contained 257 services, versus 230 in 2026-W33. x402 accounted for 255 of those 257; A2A-x402 and AP2 each accounted for one. This is less evidence of settled protocol plurality than of an early market choosing the implementation path with the most readily observable payment handshake.

Where settlement is accumulating

The income distribution is exceptionally concentrated. Across 2,020 earning agents and 1,004 earning wallets, the median 30-day observed settlement was $0.275 and the p95 was $44.5527. The Gini coefficient was 0.994, while the Herfindahl-Hirschman Index rose from 3,677.40 to 5,026.50, an increase of 1,349.10 (+36.7%) week over week. In other words, new earners are appearing, but observed value capture is becoming more—not less—centralised.

That pattern is consistent with the market’s current architecture. The largest payment rail by recorded protocol count is x402, at 3,557, versus 83 for MPP, 33 for A2A-x402, 20 for AP2, four for L402 and one for Stripe SPT. Coinbase’s March update broadened x402 beyond EIP-3009 to support ERC-20 assets using Permit2 and gas-sponsorship extensions, lowering a practical integration barrier for services that do not settle in a small set of supported stablecoins. Coinbase (coinbase.com)

Demand discovery offers one concrete opportunity area: audio transcription showed 160 variants across 501 unique sources, with an unmet share of 53% and a rising trend of 18514.3%. That is a more actionable signal than generic “agent marketplace” supply growth: it indicates a repeated, identifiable intent where discovery demand is outpacing available fulfilment.

Protocol and infrastructure: implementation is moving to the edge

The protocol stack is maturing unevenly. The A2A JavaScript SDK promoted its v1.0 line to stable on July 22, 2026, covering JSON-RPC, HTTP+JSON/REST and gRPC while retaining opt-in v0.3 compatibility; a July 28 patch addressed legacy-card normalization and signing canonicalisation. A2A JavaScript SDK changelog (github.com) Model Context Protocol’s July 28 revision remains an RC, with the specification repository explicitly warning implementers that changes may still arrive before final release and that version negotiation will be necessary during adoption. MCP releases (github.com)

The practical commercial development is Cloudflare’s push to put settlement directly in the request path. Its Monetization Gateway, announced July 1, is designed to charge for pages, datasets, APIs and MCP tools over x402, while Cloudflare Wallets, announced August 4, adds programmable wallet controls such as allowances, allow lists and maximum transaction sizes. Cloudflare Monetization Gateway (blog.cloudflare.com) Cloudflare Wallets (blog.cloudflare.com) The implication is that endpoint monetisation may move from bespoke billing integrations toward edge-enforced, policy-defined payment challenges.

AP2 is moving on a separate governance track. Google donated the protocol to the FIDO Alliance in April and released AP2 v0.2 with support for “Human Not Present” transactions, placing autonomous purchases under pre-authorised user instructions. Google Pay (blog.google) Agenstry’s 20 AP2-labelled services remain small beside x402, but the design focus on delegated authority and accountability addresses a gap that HTTP-native payment transport alone does not resolve.

Security and regulatory pressure are becoming product requirements

The strongest research warning this week is not about model capability but payment execution. A July arXiv study of 15 x402 facilitators reports violations of its proposed security rules across every evaluated facilitator, identifying attacks involving unpaid service, asset theft, denial of service and sponsor-paid gas abuse; the authors say issues were disclosed and mitigations adopted, including changes by Coinbase. “When HTTP 402 Meets the Blockchain” (arxiv.org) This makes signed requests, resource-bound authorisations, replay controls and observable settlement verification immediate procurement questions.

Policy attention is also broadening from model outputs to market conduct. On August 19, the FTC sought comment on a proposed enforcement policy regarding personalised pricing, defined as using personal data to set individual consumers’ prices. FTC policy statements (ftc.gov) For agentic commerce, the relevant issue is whether an agent can demonstrate the principal’s authority, the basis of a quoted price and the constraints applied to an automated purchase.

What to watch next week

Readers should watch whether new x402 endpoints shift from declarations to live challenge evidence, particularly as edge-native monetisation and wallet tooling become available. The more consequential metric will not be the next rise in indexed cards, but whether signed, live endpoints and independently observable settlement begin to grow faster than listings.

Sources

  1. The Next Evolution of x402: ERC-20 Support | Coinbase
  2. a2a-js/CHANGELOG.md at main · a2aproject/a2a-js · GitHub
  3. Releases · modelcontextprotocol/modelcontextprotocol · GitHub
  4. Announcing the Monetization Gateway: charge for any resource behind Cloudflare via x402 | Cloudflare Blog
  5. Announcing Cloudflare Wallets: The programmable wallet for the agentic Internet | Cloudflare Blog
  6. Google donates Agent Payments Protocol to FIDO Alliance
  7. When HTTP 402 Meets the Blockchain: Risks on Emerging x402 Payments
  8. Policy Statements | Federal Trade Commission
  9. Coinbase
  10. A2A JavaScript SDK changelog
  11. MCP releases
  12. Cloudflare Monetization Gateway
  13. Cloudflare Wallets
  14. Google Pay
  15. “When HTTP 402 Meets the Blockchain”
  16. FTC policy statements
1 · Supply funnel

Most of the agent economy is indexed, not yet operational.

Counting agents conflates supply with operation. Below is the corpus broken into five stages: what fraction of indexed agents pass each gate. The decay curve is the honest read on how mature the agent web is at any given moment.

Stage Count % of indexed Definition
Indexed 7,047
100.0%
Every candidate we have seen across every source.
Valid card 5,513
78.2%
Returned a parseable A2A/MCP card on probe.
Live 546
7.7%
Live JSON-RPC / introspectable MCP.
Signed 71
1.0%
Card published with a JWS signature.
Earning (30d) 2,020
28.7%
Agents where we observed inbound USDC/EURC settlement to their declared x402 payTo wallet. Measured over the wallets scanned so far (1474 of 1544 with a wallet) — not an ecosystem-wide total.
2 · Live-rate by ecosystem

Which source registries actually ship live agents.

Each row shows one source registry (Anthropic's MCP directory, Smithery, Glama, Postman, GitHub well-knowns, Wayback CDX scans) alongside how many of the agents it lists actually respond live on probe. The gap between seen-count and live-count is the source's effective freshness.

Source Seen Live Live rate
recrawl_hot 3,208 396 12.3%
agentic_market 2,844 76 2.7%
recrawl_warm 2,530 170 6.7%
mcp_registry 1,714 246 14.4%
cdp_discovery 1,332 41 3.1%
manifests 885 92 10.4%
github_code 772 66 8.5%
lists 482 20 4.1%
x402_list 396 24 6.1%
registry 348 190 54.6%
recrawl_cold 290 2 0.7%
n8n_workflows 194 0 0.0%
github_topics 150 4 2.7%
a2a_registry_org 116 39 33.6%
a2aregistry 114 87 76.3%
3a · By protocol kind

A2A vs MCP vs paid_api

paid_api
3,545
a2a_agent
1,964
a2a
4
3b · By observed payment-protocol label

Index composition, not global market share. x402 has multiple dedicated catalogs and a live HTTP probe; AP2, L402 and Stripe ACP are predominantly visible only when cards declare them.

x402 / AP2 / Stripe MPP / L402

x402
3,557
mpp
83
a2a_x402
33
ap2
20
l402
4
stripe_spt
1
3b · Evidence grade

How we know they take payment

314 of 3,698 payment-labelled agents (8.5%) proved it themselves — their own endpoint answered an unpaid request with an HTTP 402 challenge, or the host serves a /.well-known/x402 manifest. The rest is declared in a static agent card or described by a third-party catalog.

Facilitator feed
2,055
Catalog listing
962
Agent card
262
Live 402 challenge
232
Official protocol directory
82
/.well-known/x402
76
Not re-probed yet
23
mpp_challenge
6
3c · Indexed protocol labels over time · last 12 weeks

Newly indexed agents per week, by payment label

Changes reflect both ecosystem activity and changes in crawler source coverage.

2026-W23 · x402: 7 23 2026-W24 · ap2: 1 2026-W24 · l402: 1 2026-W24 · x402: 7 2026-W25 · a2a_x402: 1 2026-W25 · ap2: 2 2026-W25 · x402: 17 25 2026-W26 · a2a_x402: 1 2026-W26 · ap2: 1 2026-W26 · x402: 5 2026-W27 · a2a_x402: 4 2026-W27 · x402: 9 27 2026-W28 · a2a_x402: 4 2026-W28 · l402: 1 2026-W28 · x402: 1141 2026-W29 · a2a_x402: 2 2026-W29 · ap2: 1 2026-W29 · mpp: 1 2026-W29 · x402: 243 29 2026-W30 · a2a_x402: 2 2026-W30 · x402: 301 2026-W31 · a2a_x402: 1 2026-W31 · ap2: 1 2026-W31 · mpp: 81 2026-W31 · x402: 261 31 2026-W32 · a2a_x402: 8 2026-W32 · ap2: 2 2026-W32 · x402: 219 2026-W33 · a2a_x402: 4 2026-W33 · ap2: 3 2026-W33 · x402: 223 33 2026-W34 · a2a_x402: 1 2026-W34 · ap2: 1 2026-W34 · x402: 255 34
a2a_x402 ap2 l402 mpp x402
4 · Inflow concentration · 30d

A few settlement wallets receive most of the money.

Across 1,004 earning wallets backing 2,020 linked services, the distribution is heavily power-law. Wallet-level Gini and HHI avoid multiplying transfers when several catalog entries share one recipient.

Observed inflow · 30d
$256852.27
Recipient wallets
1,004
Median · wallet
$0.275
p95 · wallet
$44.55
Top 1 share
69.6%
Top 5 share
93.8%
Gini
0.994
extreme inequality
HHI
5026
highly concentrated
5 · 30-day growth

New agents arriving daily, by first-seen.

Daily count of agents we discovered for the first time. Sourced from every registry + open-web crawl we run.

2026-07-26 peak 331/day 2026-08-24
New agents · 30d
+1,850
Earning agents · 30d
2,031
observed in the snapshot table
6 · Demand gaps

What the market asked for and didn't get.

Every section above measures the supply we indexed. This one measures demand we couldn't serve: clustered search intents — from browsers, AI assistants and agents calling our A2A and MCP endpoints — where the index returned nothing or almost nothing. Aggregate only, and shown only for intents searched by enough distinct sources to be a market signal rather than one person's session.

Intent Demand Unmet Trend
audio transcription very active 53% rising

Full list + method: what the agent economy is missing. Building one of these? Add pay-per-call payments to your API →

Cite this report

Public domain numbers (CC BY 4.0). Pick the format that fits. The page auto-updates so the URL is the canonical pointer; the date stamps the snapshot you cited.

Plain text
Agenstry Research. "State of the Agent Economy." Accessed 2026-08-24. https://agenstry.com/reports/state-of-agent-economy
BibTeX
@misc{ '{' }}agenstry_state_of_agent_economy,
  title  = {State of the Agent Economy},
  author = {Agenstry Research},
  url    = {https://agenstry.com/reports/state-of-agent-economy},
  note   = {Accessed 2026-08-24},
  year   = {2026}
}
Embed (iframe)
<iframe src="https://agenstry.com/reports/state-of-agent-economy" width="100%" height="900" frameborder="0"></iframe>
Scope What this report measures, and what it doesn't

All inflow numbers are on-chain: direct eth_getLogs scans of USDC Transfer events (plus EURC on Base + Ethereum) into each indexed agent's payment_wallet across Base, Ethereum, Polygon, Arbitrum, Optimism (EVM via eth_getLogs) and Solana (SPL via getSignaturesForAddress). Reproducible from public RPC; not based on self-reporting. We report money arriving at those wallets — we do not verify that each transfer was an arm's-length sale, and we do not net out transfers an operator sends itself. See the measurement boundary.

Not included: money agents take via Stripe (per-agent Stripe accounts are private), Patreon / Sponsors, direct credit cards, PayPal, or any off-chain rail. Agenstry's own platform-skill revenue (compose / agent_stats / etc.) is also excluded: it lives in a separate accounting table that never feeds public totals. This is "the observable on-chain slice of the agent economy", not "all agent revenue ever". Off-chain rails will appear here only when the operator opts in to a future verified-reporting feed.

Methodology

Every figure on this page is computed from the same observability tables that power the rest of the Agenstry surface, with no synthetic data and no manual curation. The crawl ingests eight federated sources plus open-web well-known probes. Agents are scored against a 9-criterion conformance methodology. Observed inflow is derived from on-chain USDC/EURC transfer scans across six chains: Base, Ethereum, Polygon, Arbitrum, Optimism, and Solana; AP2 / Stripe MPP / L402 receipts are detected on agent cards but indexed only when the rail is publicly verifiable. The full crawl + scoring pipeline is open behind the federation feed. If our numbers disagree with yours, one of us has a bug, and we would like to know.