What differs between the 27 August 2026 and 11 September 2026 readings
Two dated readings of the Circle x402 Discovery catalog, compared. Composition only.
⭐ Where the money is pointed
27
offers name a different payout address in the later reading, out of 2,315
offers present in both.A seller repointing where its money goes is not announced anywhere, and a buyer paying from a cached listing would not notice. ⛔ This says nothing about who controls either address. A changed entry is a changed catalog entry; ownership was not checked on any chain, and routine key rotation and a takeover look identical from here.
What the pattern looks like.
Between the 2026-08-27 and 2026-09-11 readings, 27 offers name a
different payout address, and 24 of them belong to a
single host, which changed on Base and Solana mainnet in the same window.
They point to 27
distinct addresses — one per offer, none reused,
and none of them appears anywhere in the earlier reading.
Both figures are derived from the two harvests, not recorded by hand.
The ordinary explanation fits it. Deriving one payout address per endpoint is a reasonable and common design — it makes revenue attributable per route without a ledger — and a service built that way regenerates every address together when it re-keys or redeploys. A host rotating all of its endpoints at once, on every chain it serves, is exactly what that looks like. ⭐ Diversion tends the other way: someone redirecting payments wants few destinations, not 27.
What would tell the two apart, and has not been done: ask each endpoint for its own
⛔ And there is a question this instrument cannot reach at any effort. A catalog records where money is pointed. It never records who pointed it, or why. A repoint is a write to Circle's index; no chain witnessed it, so no chain can testify about it. Nothing here is evidence about any operator's conduct, and no amount of further care with this data would make it so.
The ordinary explanation fits it. Deriving one payout address per endpoint is a reasonable and common design — it makes revenue attributable per route without a ledger — and a service built that way regenerates every address together when it re-keys or redeploys. A host rotating all of its endpoints at once, on every chain it serves, is exactly what that looks like. ⭐ Diversion tends the other way: someone redirecting payments wants few destinations, not 27.
What would tell the two apart, and has not been done: ask each endpoint for its own
402 challenge and see whether the address it names is the one in this reading. That is
the operator's own server asserting the change, which no third party can assert for it.
Those requests have not been made — not by us, and as far as we know by nobody else.
⛔ On-chain reads would not settle it either: independent addresses are consistent with per-endpoint
derivation and with freshly generated keys, and a shared deployer is consistent with
systematic provisioning and with systematic exfiltration.
⛔ And there is a question this instrument cannot reach at any effort. A catalog records where money is pointed. It never records who pointed it, or why. A repoint is a write to Circle's index; no chain witnessed it, so no chain can testify about it. Nothing here is evidence about any operator's conduct, and no amount of further care with this data would make it so.
| resource | network | payout in 2026-08-27 | payout in 2026-09-11 | state |
|---|---|---|---|---|
| https://agents.allium.so/api/v1/developer/prices | Base | 0x62a5b26d…e18b43 | 0xfa0a25a4…b10ab4 | verified-changed |
| https://agents.allium.so/api/v1/developer/prices | Solana mainnet | F2xY6baQKh…KRNRBL | EyvpoXHQkM…MjbWdg | verified-changed |
| https://agents.allium.so/api/v1/developer/prices/at-timestamp | Base | 0xe46dcc08…958e32 | 0x89df5d20…988580 | verified-changed |
| https://agents.allium.so/api/v1/developer/prices/at-timestamp | Solana mainnet | 7FgXSKPkmP…TJNAmu | JqtThqWtfU…rhRyVn | verified-changed |
| https://agents.allium.so/api/v1/developer/prices/history | Base | 0x3725f70a…8e08e8 | 0xbf677642…89b124 | verified-changed |
| https://agents.allium.so/api/v1/developer/prices/history | Solana mainnet | E41JahfCc7…mfHnJM | 4yNA9WzLae…LuVpid | verified-changed |
| https://agents.allium.so/api/v1/developer/prices/stats | Base | 0x09d7a7a1…1d11a0 | 0xd3ecace3…405543 | verified-changed |
| https://agents.allium.so/api/v1/developer/prices/stats | Solana mainnet | E4g1Frgpqg…1Hxwaf | AeYcFSA2oQ…e2kpL8 | verified-changed |
| https://agents.allium.so/api/v1/developer/tokens | Base | 0x91b016ec…948b40 | 0xe92887a7…f9adf0 | verified-changed |
| https://agents.allium.so/api/v1/developer/tokens | Solana mainnet | EDdYLUVrVb…AhKFr6 | 7fKsgnFCUk…USVRBb | verified-changed |
| https://agents.allium.so/api/v1/developer/tokens/chain-address | Base | 0x835898a0…933034 | 0x2b8b7dff…a119ec | verified-changed |
| https://agents.allium.so/api/v1/developer/tokens/chain-address | Solana mainnet | AD3T6bUxgc…YEBC28 | JC7HmLdXz5…u6m1bp | verified-changed |
| https://agents.allium.so/api/v1/developer/tokens/search | Base | 0x5e8943bf…746a83 | 0xbff53303…12dd78 | verified-changed |
| https://agents.allium.so/api/v1/developer/tokens/search | Solana mainnet | 5mN1FxkA5V…Vmwape | EkwtpqoqZ7…yxcG39 | verified-changed |
| https://agents.allium.so/api/v1/developer/wallet/balances | Base | 0x42dc898b…a3d853 | 0x4da95a60…739996 | verified-changed |
| https://agents.allium.so/api/v1/developer/wallet/balances | Solana mainnet | FcGxUxN9f8…YxgGpF | 3oXpBVUoBf…GaAjvd | verified-changed |
| https://agents.allium.so/api/v1/developer/wallet/balances/history | Base | 0x9dfaf23e…111d1f | 0x4a3a99e8…173942 | verified-changed |
| https://agents.allium.so/api/v1/developer/wallet/balances/history | Solana mainnet | 66HnQrPP9v…7bQ996 | DknwUUnTMV…t28exf | verified-changed |
| https://agents.allium.so/api/v1/developer/wallet/pnl | Base | 0x78654c72…acb1f6 | 0xb01c407a…5288ab | verified-changed |
| https://agents.allium.so/api/v1/developer/wallet/pnl | Solana mainnet | Fo7gGp3Pad…bCq6kL | 9ZWJodn7Ea…2mcBNN | verified-changed |
| https://agents.allium.so/api/v1/developer/wallet/transactions | Base | 0x64ec6dba…026686 | 0xed5ef79c…e3da68 | verified-changed |
| https://agents.allium.so/api/v1/developer/wallet/transactions | Solana mainnet | Axe3CL3zFC…bRaK8P | 7YUpKWNEmH…kZAyMB | verified-changed |
| https://agents.allium.so/api/v1/explorer/queries/run-async | Base | 0x70458a7b…53f779 | 0xaee41349…62a860 | verified-changed |
| https://agents.allium.so/api/v1/explorer/queries/run-async | Solana mainnet | 399cyRm6LQ…54hJNy | GTsjbmAvin…CYvqDE | verified-changed |
| https://parallelmpp.dev/api/extract | Base | 0x1673150a…a1a98a | 0x993814d2…4fa326 | verified-changed |
| https://parallelmpp.dev/api/search | Base | 0x8b0f2e46…65d7d2 | 0xcf91fc57…db317f | verified-changed |
| https://parallelmpp.dev/api/task | Base | 0x4d576268…b77b96 | 0xf9daf8bc…7d4ad9 | verified-changed |
Addresses truncated here; the JSON carries them in full. per (resource, network, asset, scheme), the SET of payTo values in each reading, compared only where both readings carry at least one. Set-valued because the key is not unique — 834 keys carry conflicting rows — so no row is paired with a specific successor.
⛔ Two instants, not a period — "unchanged" is not "stable"
These are two instants 15 days apart, not a period of observation. Nothing here reports what happened in between: a value identical on both dates may have changed and changed back, a listing present on both may have been absent for a fortnight, and a single reading on either end could itself be atypical. "Unchanged" here means "equal at two instants", which is not the same as stable.
This is stated here rather than in a footnote because every number on this page is read as a rate of change by default, and none of them is one.
⛔ What "absent" does and does not mean
Neither reading can bound what it did not see. The 2026-08-27 harvest never looked for a
So every count below says absent from the 2026-09-11 reading or present only in the 2026-09-11 reading. Neither is an event. Nobody observed a change; two readings simply differ.
total; the 2026-09-11 harvest did find one and matched it, but a listing absent from a
later reading may have been withdrawn, renamed, or missed by paging — and nothing here
distinguishes those.So every count below says absent from the 2026-09-11 reading or present only in the 2026-09-11 reading. Neither is an event. Nobody observed a change; two readings simply differ.
Presence
| 2026-08-27 | 2026-09-11 | in both | absent from 2026-09-11 | only in 2026-09-11 | |
|---|---|---|---|---|---|
| distinct resource URLs | 960 | 1,094 | 877 | 83 | 217 |
| distinct payout addresses | 111 | 128 | 83 | 28 | 45 |
| hosts | 25 | 48 | 25 | 0 | 23 |
| accepts rows | 3,808 | 4,215 | — | — | — |
Hosts present only in the 2026-09-11 reading
2captcha.x402.paysponge.com · agentres.dev · api.paysponge.com · api.socialintel.dev · api.syntalic.com · api.venice.ai · app.earnfi.fun · bigquery.google.gateway-402.com · embeddings.alibaba.gateway-402.com · fal.x402.paysponge.com · generativelanguage.google.gateway-402.com · mcp.blocksize.info · modelstudio.alibaba.gateway-402.com · ocr.alibaba.gateway-402.com · public-api.birdeye.so · rentcast.x402.paysponge.com · screenshotone.x402.paysponge.com · texttospeech.google.gateway-402.com · translate.alibaba.gateway-402.com · travel.brij.fi · tripadvisor.x402.paysponge.com · wolframalpha.x402.paysponge.com · x402-api.vybenetwork.xyz
No host present in the 2026-08-27 reading is absent from the 2026-09-11 one. Every difference in the host column is one direction only.
Advertised price
208 offers advertise a different amount in the later
reading.
⛔ Advertised, not charged. No endpoint was called in either reading, so neither says what any of these services actually asks for when paid.
⛔ Advertised, not charged. No endpoint was called in either reading, so neither says what any of these services actually asks for when paid.
| resource | network | 2026-08-27 | 2026-09-11 |
|---|---|---|---|
| https://api.aisa.one/apis/v2/coingecko/coins/categories | Ethereum | 8000 | 12000 |
| https://api.aisa.one/apis/v2/coingecko/coins/categories | Base | 8000 | 12000 |
| https://api.aisa.one/apis/v2/coingecko/coins/categories | Avalanche | 8000 | 12000 |
| https://api.aisa.one/apis/v2/coingecko/coins/categories | Arbitrum | 8000 | 12000 |
| https://api.aisa.one/apis/v2/coingecko/coins/categories | Optimism | 8000 | 12000 |
| https://api.aisa.one/apis/v2/coingecko/coins/categories | Polygon | 8000 | 12000 |
| https://api.aisa.one/apis/v2/coingecko/coins/categories | Unichain | 8000 | 12000 |
| https://api.aisa.one/apis/v2/coingecko/coins/categories/list | Ethereum | 8000 | 12000 |
| https://api.aisa.one/apis/v2/coingecko/coins/categories/list | Base | 8000 | 12000 |
| https://api.aisa.one/apis/v2/coingecko/coins/categories/list | Avalanche | 8000 | 12000 |
| https://api.aisa.one/apis/v2/coingecko/coins/categories/list | Arbitrum | 8000 | 12000 |
| https://api.aisa.one/apis/v2/coingecko/coins/categories/list | Optimism | 8000 | 12000 |
First 12 of 208; the JSON carries the method.
Networks
| network | CAIP-2 | 2026-08-27 | 2026-09-11 | difference |
|---|---|---|---|---|
| Base | eip155:8453 | 1,164 | 1,327 | +163 |
| Solana mainnet | solana:5eykt4UsFv8P8NJdTREpY1vzqKqZKvdp | 562 | 823 | +261 |
| Polygon | eip155:137 | 625 | 555 | -70 |
| Base Sepolia | eip155:84532 | 396 | 396 | 0 |
| Polygon Amoy | eip155:80002 | 396 | 396 | 0 |
| Solana devnet | solana:EtWTRABZaYq6iMfeYKouRu166VU2xqa1 | 264 | 264 | 0 |
| Avalanche | eip155:43114 | 98 | 108 | +10 |
| Arbitrum | eip155:42161 | 56 | 69 | +13 |
| Ethereum | eip155:1 | 56 | 66 | +10 |
| Optimism | eip155:10 | 56 | 66 | +10 |
| Unichain | eip155:130 | 56 | 66 | +10 |
| X Layer | eip155:196 | 31 | 31 | 0 |
| Sonic | eip155:146 | 12 | 12 | 0 |
| World Chain | eip155:480 | 12 | 12 | 0 |
| Sei | eip155:1329 | 12 | 12 | 0 |
| HyperEVM | eip155:999 | 12 | 12 | 0 |
The gate this page had to pass first
| reading | rows | Base Sepolia | Polygon Amoy | verdict |
|---|---|---|---|---|
| 2026-08-27 | 3,808 | 396 | 396 | passed |
| 2026-09-11 | 4,215 | 396 | 396 | passed |
Both readings must carry testnet rows or this page renders a refusal instead of a comparison — a filtered reading has a different denominator, and the difference would appear here as roughly 165 testnet listings vanishing. Counted from the rows, not from either harvest's stored self-check.
What this comparison does NOT claim
| field | state | why |
|---|---|---|
| whatHappenedBetween | not-measured | two readings bound their endpoints and say nothing about the interval; a value equal on both dates may have changed and changed back |
| causeOfAbsence | cannot-distinguish | a listing absent from the later reading may be withdrawn, renamed, or missed by paging — neither harvest can bound what it did not see |
| payoutOwnership | not-measured | a changed payTo is a changed catalog entry; whether the same party controls both addresses was not checked on any chain |
| whoChangedTheEntry | structurally-unanswerable | the catalog records where money is pointed, never who pointed it or why. A repoint is a write to Circle's index; no on-chain read can testify about it, so this is a limit of the instrument rather than a gap in the effort |
| liveness | not-measured | no third-party endpoint was called in either reading, so neither says whether any of these services answered |
Method
Both readings were fetched directly from api.circle.com/v2/x402/discovery/resources
with limit=100 and no siwx parameter. This page compares them; it makes no
request of its own and never has.
⛔ The join is set-valued, not row-paired. (resource, network, asset, scheme)
is not unique — 834 keys in the later reading carry rows with conflicting values, the same URL
offered at two prices — so pairing rows would silently drop a third of them. Per key this compares
the SET of payout addresses and the SET of amounts, which invents no identity and loses no row.
The machine-readable form, with full addresses: /snapshot/diff/2026-08-27..2026-09-11.json