— PAID DEMAND
Which AI services have actually been paid.
Directories list thousands of x402 services and publish no evidence about any of them. We asked the chain about every payment address we could resolve.
997 of 1,079 have been paid at least once — 82 never have. Together: $247,260 across 10,414,904 payments. The median service has earned $0.51.
Read from a subgraph on The Graph Network, through The Graph’s Subgraph MCP server. The call is printed below — re-run it yourself. Free, CC BY 4.0.
The shape of it
- $247,260
- settled, all time across 10,414,904 payments
- $0.51
- the median service 592 have earned under $1
- 67%
- is one receiver the top five are 89%
- 82
- listed but never paid of 1,079 addresses we asked about
A sum is the one statistic a concentrated economy always passes. Here is the distribution instead.
Mean payment: $0.02. Fractions of a cent, at enormous volume.
| All-time earnings | Receivers | Share of the paid set |
|---|---|---|
| under $1 | 592 | 59% |
| under $10 | 827 | 83% |
| $1,000 or more | 9 | 1% |
The servers we measure that have been paid
49 of them publish a payment address of their own, so the receipts are theirs and nobody else’s.
| # | Capability | Settled | Payments | Per call | tashan score |
|---|---|---|---|---|---|
| 1 | Blockrun npm api.aidress.ai | $166,659 | 8,718,732 | $0.02 | 68 |
| 2 | scvd.store | $120 | 81 | $1.48 | 52 |
| 3 | agent402.tools · 2 payment addresses | $110 | 9,079 | $0.01 | 65 |
| 4 | Pqs npm promptqualityscore.com | $36.53 | 1,107 | $0.03 | 44 |
| 5 | Coinopai npm x402.coinopai.com | $33.59 | 690 | $0.05 | 61 |
| 6 | 2sio npm 2s.io | $28.91 | 10,107 | <$0.01 | 60 |
| 7 | Vaaya npm vaaya.ai | $17.60 | 530 | $0.03 | 66 |
| 8 | X402 Bazaar npm 402.com.tr | $11.45 | 281 | $0.04 | 50 |
| 9 | Signalfuse npm api.signalfuse.co | $6.63 | 491 | $0.01 | 57 |
| 10 | Tensorfeed npm tensorfeed.ai | $5.06 | 205 | $0.02 | 57 |
| 11 | Skim402 npm skim402.com | $4.37 | 894 | <$0.01 | 45 |
| 12 | Stelar Signals npm api.stelardigital.com | $3.90 | 25 | $0.16 | 59 |
| 13 | minia2a.uk | $3.79 | 71 | $0.05 | 59 |
| 14 | api.osf-master-server.com | $3.75 | 89 | $0.04 | 45 |
| 15 | Tenjin npm tenjin.blog · 2 payment addresses | $3.46 | 12 | $0.29 | 56 |
| 16 | agents.ai-rook.com | $3.36 | 118 | $0.03 | 46 |
| 17 | X402 Trust npm x402.fuchss.app | $3.12 | 56 | $0.06 | 56 |
| 18 | Rubric Protocol npm rubric-protocol.com | $3.02 | 177 | $0.02 | 51 |
| 19 | Arena npm agentarena.lol | $3.00 | 3 | $1.00 | 54 |
| 20 | ai-data-marketplace-1042299154756.us-central1.run.app | $2.94 | 46 | $0.06 | 50 |
| 21 | Connskill Growth npm agent.connskill.com | $2.70 | 31 | $0.09 | 49 |
| 22 | Tollbooth X402 npm x402toll.com | $2.55 | 20 | $0.13 | 45 |
| 23 | Vedetta npm vedetta.dethboy.com | $2.50 | 42 | $0.06 | 47 |
| 24 | x402.forgemesh.io · 2 payment addresses | $1.20 | 168 | <$0.01 | 56 |
| 25 | Musicwire npm musicwire.5432wire.com | $1.15 | 6 | $0.19 | 54 |
| 26 | Fiatdock npm fiatdock.com · 5 payment addresses | $1.04 | 92 | $0.01 | 47 |
| 27 | Orcpin npm orcpin.dev | $0.99 | 33 | $0.03 | 53 |
| 28 | Toll402 npm toll402.dev | $0.96 | 66 | $0.01 | 63 |
| 29 | Agentpay npm agentpay.tools | $0.76 | 95 | <$0.01 | 52 |
| 30 | library.forgemesh.io | $0.69 | 179 | <$0.01 | 56 |
| 31 | Dex Data npm x402.donnyautomation.com | $0.54 | 54 | <$0.01 | 52 |
| 32 | Crypto Signals npm crypto.forgemesh.io | $0.51 | 13 | $0.04 | 55 |
| 33 | X402 Ads npm ads.forgemesh.io | $0.46 | 24 | $0.02 | 59 |
| 34 | X402 Seller npm x402-seller-0ay3.onrender.com | $0.45 | 63 | <$0.01 | 58 |
| 35 | Url2md IO npm url2md.io | $0.42 | 82 | <$0.01 | 57 |
| 36 | Voice npm voice.forgemesh.io | $0.29 | 100 | <$0.01 | 56 |
| 37 | Miniframe Tools npm tools.miniframe.com.br | $0.27 | 15 | $0.02 | 45 |
| 38 | Orizn Visa npm visa.orizn.app | $0.15 | 5 | $0.03 | 48 |
| 39 | Lazaretto npm lazaretto.dev | $0.12 | 4 | $0.03 | 50 |
| 40 | Groundcheck npm groundcheck.seiche.info | $0.07 | 3 | $0.02 | 46 |
| 41 | Agent Ready npm agent-ready.dev | $0.06 | 3 | $0.02 | 57 |
| 42 | Q402 npm q402.quackai.ai | $0.04 | 2 | $0.02 | 68 |
| 43 | DNS Doctor npm dnsdoctor.dev | $0.03 | 4 | <$0.01 | 48 |
| 44 | X402 Notary npm notary.forgemesh.io | $0.03 | 9 | <$0.01 | 59 |
| 45 | Assay npm assay.nominal-labs.com | $0.02 | 4 | <$0.01 | 50 |
| 46 | Orchard Data npm orchard-data.letom1176.workers.dev | $0.01 | 4 | <$0.01 | 45 |
| 47 | Bykaranteli npm bykaranteli.com | $0.01 | 4 | <$0.01 | 60 |
| 48 | Spoolis npm spoolis.com | <$0.01 | 3 | <$0.01 | 56 |
| 49 | Tashan CLI npm tashan.sh | $0.00 | never paid | — | 55 |
Payment is not an input to the tashan score. “Someone pays for this” and “this is well made” are different claims — the top earner here scores 71.
We are in this table. Tashan CLI publishes a price and has settled $0.00, ever. An instrument that exempts itself from its own measurement is not an instrument.
What this does not mean
- Charging is not being paid. A service can publish a price and never settle a single call. Both facts are on this page and they are different columns.
- Unpaid is not bad. Almost every MCP server is free by design; an absent row here says nothing about quality. It says nobody has paid this address, because nobody was asked to.
- Volume belongs to an address, not to a project. Where one payment address is shared by several services the chain cannot separate them, so those receipts are counted in the totals above and never attributed to a capability. 11 matched receivers are excluded from the table for exactly that reason.
- This is Base, and x402 only. Other chains, other payment rails and private billing are invisible here. We publish what is on the ledger we can read.
Re-run it yourself
Two Graph products, composed: a subgraph on The Graph Network, read through The Graph’s Subgraph MCP server. Point any MCP client at it and you get the same rows.
https://subgraphs.mcp.thegraph.com/sse
tool: execute_query_by_subgraph_id
subgraph_id: Cb56epg3EvQ6JRpPfknbkM54QxpzTvLa7mwKNQQfUyoj
{ x402AddressSummaries(first: 1000, where: {role: RECIPIENT}) {
address totalPayments totalVolume } }From a clone of the repo, one command runs exactly that hop:
GRAPH_API_KEY=<your key> python3 pipeline/subgraph_mcp.py --probeAddresses from the Bazaar’s own listings. Receipts from the chain. The join is an exact host match against a homepage or repository we already hold — never a fuzzy name match.
Source: subgraph Cb56epg3EvQ6JRpPfknbkM54QxpzTvLa7mwKNQQfUyoj via subgraphs.mcp.thegraph.com. Recomputed nightly.