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 2026-11-18 23:14:15 UTC methodology JSON CSV cite this archive
Weekly briefing · 2026-W25 archived view
Jun 15, 2026

Agentic AI deployment remains heavily concentrated and growth is modest. This week the index of known agents rose 2.7% to 2,998, with 187 live endpoints (JSON-RPC probes) – a 14.7% jump. Yet discovery has largely stalled: only 582 new agents appeared over the past 30 days (down 80.1% from 2,919) and still only 21 agents generated any revenue. Those 21 earned a total of $491.02 (30-day gross), with a median haul of $6.31 (95th percentile $99.371) – the top five agents captured 79.2% of revenue (top one 21.7%), yielding a Gini of 0.683. In short, a tiny minority of agents now dominate an ecosystem barely inching forward.

Where the money is flowing

The financial upside of the agent economy is still negligible but showing early signs of formal structure. On-chain revenues reported through the x402/Stripe protocols totaled just $491.02, implying token payouts in the dollars–tens. This week’s data underscores that agent payments remain microscopic: only 21 agents earned, with median $6.31 per agent over 30 days. By comparison, global stablecoin activity is astronomic: Andreessen Horowitz’s State of Crypto reckons stablecoins cleared $46 trillion in transaction volume over the last year (www.lemonde.fr) – orders of magnitude above Visa/PayPal. Stablecoins and crypto rails (notably x402) are effectively the payment backbone for agents. Silicon-valley and payments incumbents are investing heavily. Jeff Bezos’s industrial-AI startup Prometheus just raised $12 billion at a $41 billion valuation (techcrunch.com) – evidence that large investors see big things in “AI agents” (even if Prometheus targets robotics). Meanwhile, Stripe and Google have doubled down on agentic commerce, and Circle Internet (issuer of USD Coin) launched a new “Agent Stack” product suite (CLI, agent wallets, marketplace, nanopayments) to let AI agents hold funds and transact autonomously (www.circle.com). All of these moves hinge on crypto rails and stablecoins. Small wonder that industry observers say stablecoins have become the “backbone of a new global payment system” (www.lemonde.fr). We chart payment-rail usage below: x402 dominates (113 endpoints) versus trivial counts for legacy/mocks (4 l402, 1 AP2).

Agent payment rails used118x402 (HTTP 402)113 · 95.8%L402 (legacy)4.00 · 3.4%AP2 (Google)1.00 · 0.8%

The donut chart shows the payment-protocol breakdown of discovered agent endpoints: x402 is by far the dominant rail (113 agents) with AP2 and legacy L402 nearly absent. This reflects recent industry trends (coinbase’s “neutral” x402 standard under Linux Foundation involves Amazon, Google, Visa, etc (www.theblock.co), whereas Google’s earlier AP2 protocol has only 1 live agent here).

Spec & protocol

Standards work continues apace. Google’s A2A protocol (agent-to-agent communication) is now established under the Linux Foundation; Google reports “more than 150 organizations” backing A2A in production use (www.linuxfoundation.org). The upstart MCP (Model Context Protocol) is on schedule for a big mid-year update: a release candidate for the July-28, 2026 spec arrived May 21, featuring a new stateless core, extensions framework, and first-class “Tasks” (blog.modelcontextprotocol.io). Implementers are advised to start testing migration (e.g. eliminating sessions and legacy handshake) but to await the final 2026-07-28 release (blog.modelcontextprotocol.io). In payments, Google formally donated its Agent Payments Protocol (AP2) to the FIDO Alliance; FIDO has launched a new “Agentic Authentication” working group aiming to develop open standards for agentic commerce (building on AP2 and Mastercard’s Verifiable Intent) (fidoalliance.org). At the same time, the coalition around Linux Foundation’s x402 rail is broadening to include Visa, Mastercard, Microsoft, Shopify, Solana Labs, and others (www.theblock.co) – a sign that industry payments giants are betting on decentralized micropayment rails.

Notable launches

On the products side, new agent platforms and tools continue to appear. In financial infrastructure, Circle’s “Agent Stack” (May 2026) provides wallets, marketplaces, and CLI tools so AI agents can store assets and pay autonomously (www.circle.com). Enterprises are also formalizing agent development. Glean (enterprise search) introduced an “Enterprise Agent Development Lifecycle” at Google I/O, setting a framework and governance model for how companies should build and measure AI agents (www.glean.com). In the broader AI-tool race, Kore.ai revealed its new Artemis platform (May 2026) – a managed agent-builder service that includes a declarative “Agent Blueprint Language” and native support for Microsoft’s agent infrastructure. This lets enterprises “build, govern, and optimize AI agents using AI itself” (the platform even uses machine learning to accelerate agent authoring) (venturebeat.com). Artemis initially runs on Azure and integrates with Microsoft’s Agent 365 suite. Notably, Kore.ai pitched Artemis as a “neutral” layer in the fragmenting agent-platform market (competing with Salesforce, Google, ServiceNow, etc.) (venturebeat.com).

Research

Academic and industry research on multi-agent AI is active. Recent preprints highlight collaborative agent capabilities. For example, one ArXiv paper shows that teams of large-language-model agents (each acting as an expert) can generate more creative solutions than even human groups. Across six brainstorming tasks, multi-agent LLM teams outperformed human teams by a large margin (Cohen’s d=1.50) (arxiv.org), driven by more novel idea generation. The authors note that wide semantic exploration (low coherence) helps agent-team creativity. Another new study introduced EinsteinArena, a web platform where autonomous AI agents collaborate on open math problems. By May 2026, agents on EinsteinArena had “discovered 12 new state-of-the-art results” – such as improving the “kissing number” problem in 11 dimensions from 593 to 604 (arxiv.org). In other words, an agent collective on EinsteinArena independently pushed a 40-year-old research record higher. These examples suggest that agentic AI is not just a fiction but already solving novel problems through multi-agent teamwork.

Regulation & policy

Regulators are still catching up on agentic AI. One focal point is payments regulation: the flood of stablecoin use for agent commerce is drawing scrutiny (stablecoins are now used orders of magnitude more than credit cards (www.lemonde.fr)). In April, the FIDO Alliance’s involvement (see above) signaled industry moving toward open standards rather than ad-hoc solutions. On the policy front, Europe is mulling tighter AI oversight. The EU’s draft “High-Risk AI” guidelines (part of the AI Act framework) are open for public comment through 23 June (www.itpro.com). That deadline (next week) is worth watching: the final guidelines will clarify what counts as a high-risk autonomous system. If autonomous commerce or agentic decision-making are eventually cited, it could usher formal regulation of AI agents as a special category.

What to watch next week

The next milestone is the close of the EU’s AI guidelines consultation on June 23 (www.itpro.com). How policymakers respond could presage formal rules for AI agents. Concurrently, industry participants will be eyeing upcoming protocol and deployment milestones: for instance, the Model Context Protocol’s full release in late July and any FIDO Alliance publications on agentic payments. Investors and engineers alike will also look for how quickly the next wave of agents can move from research platforms (like EinsteinArena) into real-world use. Given the still-small number of revenue-generating agents, any major new pilot (perhaps from a major cloud vendor or payments company) could shift the needle.

Sources

  1. 'France has six months to catch the next industrial wave of agentic AI'
  2. Jeff Bezos's Prometheus raises $12B to build an 'artificial general engineer' for the physical world | TechCrunch
  3. Circle Powers the Agentic Economy with New AI Infrastructure | Circle
  4. Tech, crypto giants to help steward Coinbase's neutral x402 payments protocol under Linux Foundation | The Block
  5. A2A Protocol Surpasses 150 Organizations, Lands in Major Cloud Platforms, and Sees Enterprise Production Use in First Year
  6. The 2026-07-28 MCP Specification Release Candidate | Model Context Protocol Blog
  7. FIDO Alliance to Develop Standards for Trusted AI Agent Interactions | FIDO Alliance
  8. Glean Introduces the Enterprise Agent Development Lifecycle, Codifying How Enterprises Build, Govern, and Measure AI Agents | Glean Press
  9. Kore.ai launches Artemis AI agent platform, takes on Salesforce and ServiceNow | VentureBeat
  10. Kore.ai launches Artemis AI agent platform, takes on Salesforce and ServiceNow | VentureBeat
  11. Multi-agent AI systems outperform human teams in creativity
  12. Harnessing the Collective Intelligence of AI Agents in the Wild for New Discoveries
  13. European Commission opens public consultation on long-awaited draft for high-risk AI guidelines
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 11,857
100.0%
Every listing we hold across every source; 4,060 archived listings are not counted here or in any step below.
Valid card 10,254
86.5%
Returned a parseable A2A/MCP card on probe.
Live 1,765
14.9%
Live JSON-RPC / introspectable MCP.
Signed 283
2.4%
Card published with a JWS signature.
Earning (30d) 4,676
39.4%
Listings whose declared payTo wallet received inbound USDC in the last 30 days (2,154 distinct wallets; several listings can share one). Measured over the wallets scanned so far (3,127 of 3,359 with a wallet) — not an ecosystem-wide total. Not a subset of Live: most are paid APIs, which answer 402 rather than a free JSON-RPC probe.
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
agentic_market 6,231 314 5.0%
cdp_discovery 6,093 408 6.7%
manifests 2,456 594 24.2%
mcp_registry 1,747 273 15.6%
n8n_workflows 1,605 0 0.0%
x402_list 1,550 211 13.6%
registry 1,094 581 53.1%
smithery 1,080 263 24.4%
a2aregistry 934 546 58.5%
github_code 906 80 8.8%
a2a_registry_org 860 447 52.0%
lists 584 85 14.6%
8004scan 247 66 26.7%
dexter_discovery 229 3 1.3%
github_topics 166 9 5.4%
3a · By protocol kind

A2A vs MCP vs paid_api

paid_api
6,705
a2a_agent
3,549
workflow
1,603
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
6,981
mpp
140
a2a_x402
108
ap2
42
l402
6
stripe_acp
1
3b · Evidence grade

How we know they take payment

3,758 of 7,278 payment-labelled agents (51.6%) 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.

Live 402 challenge
3,105
Facilitator feed
2,403
Agent card
754
/.well-known/x402
504
Catalog listing
352
mpp_manifest
111
mpp_challenge
38
Official protocol directory
11
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-W35 · a2a_x402: 3 2026-W35 · ap2: 1 2026-W35 · mpp: 1 2026-W35 · x402: 75 35 2026-W36 · a2a_x402: 3 2026-W36 · mpp: 1 2026-W36 · stripe_acp: 1 2026-W36 · x402: 310 2026-W37 · a2a_x402: 5 2026-W37 · mpp: 1 2026-W37 · x402: 251 37 2026-W38 · a2a_x402: 2 2026-W38 · ap2: 1 2026-W38 · mpp: 7 2026-W38 · x402: 143 2026-W39 · a2a_x402: 4 2026-W39 · ap2: 1 2026-W39 · mpp: 1 2026-W39 · x402: 232 39 2026-W40 · a2a_x402: 3 2026-W40 · ap2: 7 2026-W40 · mpp: 8 2026-W40 · x402: 449 2026-W41 · a2a_x402: 4 2026-W41 · ap2: 3 2026-W41 · mpp: 3 2026-W41 · x402: 360 41 2026-W42 · a2a_x402: 9 2026-W42 · ap2: 2 2026-W42 · mpp: 7 2026-W42 · x402: 438 2026-W43 · a2a_x402: 2 2026-W43 · ap2: 2 2026-W43 · mpp: 9 2026-W43 · x402: 546 43 2026-W44 · a2a_x402: 5 2026-W44 · ap2: 1 2026-W44 · l402: 2 2026-W44 · mpp: 5 2026-W44 · x402: 680 2026-W45 · a2a_x402: 17 2026-W45 · ap2: 3 2026-W45 · mpp: 4 2026-W45 · x402: 401 45 2026-W46 · a2a_x402: 5 2026-W46 · ap2: 1 2026-W46 · mpp: 1 2026-W46 · x402: 433 2026-W47 · a2a_x402: 6 2026-W47 · mpp: 2 2026-W47 · x402: 157 47
a2a_x402 ap2 l402 mpp stripe_acp x402
4 · Inflow concentration · 30d

A few settlement wallets receive most of the money.

Across 2,154 earning wallets backing 4,676 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
$158,886,429.59
Recipient wallets
2,154
Active sellers · 30d
1,799
Median · wallet
$0.122
p95 · wallet
$61.48
Top 1 share
99.7%
Top 5 share
99.9%
Gini
1.0
extreme inequality
HHI
9946
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.

2026-10-20 peak 363/day 2026-11-18
New agents · 30d
+3,131
New earning wallets · 30d
626
first observed earning in the last 30 days
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
aidata active 36% 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 2026-11-18. 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 2026-11-18},
  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

Inflow figures count public stablecoin transfers into wallets declared by indexed agents. They are reproducible from public chains, but a transfer may be self-funded and is not proof of a sale. See the measurement method.

Private and off-chain payments, and Agenstry's own paid-skill receipts, are excluded. Coverage varies by chain, asset and date.

Methodology

This report uses dated observations from the public index. Check source coverage, scoring rules and the public feed before comparing it with another dataset.