Historical: BNB Chain rows from before our 2026-07-28 scope change are retained for reproducibility. Current scope is Ethereum mainnet, Base, and Arbitrum One. Three of the six rows below (ethereum, base, arbitrum) are the current-scope census; the three BNB rows (bnb_standard, bnb_standard_late, bnb_brc8004_fork) are the historical slice kept intact so prior citations resolve to the same numbers.
Public ERC-8004 totals published by third-party explorers and press mentions disagree with each other by orders of magnitude in this window; this census publishes its own per-chain rows and its own method rather than adopting any of them. None of those external figures answer the question a builder or funder actually asks: how many distinct operators are behind those registrations, and on which chain do they live? This dataset is a six-row measurement of that. Same registry contract (0x8004A169FB4a3325136EB29fA0ceB6D2e539a432) on Ethereum, Base, Arbitrum One (current scope), and BNB Smart Chain (historical); four separate eth_getLogs re-scans against Alchemy archive RPCs; one checkpoint clock at 2026-09-17T04:31Z UTC (per-chain block anchors inside the table); three whole-history censuses (Ethereum, Base, Arbitrum, 100% tx.from coverage) alongside one BSC whole-history raw-count row and two disjoint BSC cluster samples reported as lower bounds with Wilson CI95s. Numbers from a census and numbers from a sample sit in separate columns so no reader slots them side by side. Seven named claim traps are enforced in the schema and re-flagged in GRADE.md — read that file before citing any single row in isolation.
https://api.onchainagentintel.io/v1/intel/agent/{token_id}?chain={chain}, using the x402 payment scheme; discovery is at /.well-known/x402.json. The offer is additive to this CC BY 4.0 dataset — the dataset itself is free to reuse; the paid path buys live per-agent detail this snapshot does not carry.
The 2026-09-17 census table (six rows — three current-scope, three historical)
Every value below is verbatim from dataset.csv.
Column order matches the CSV emission order. Empty cells are structurally
empty (e.g. Wilson CI95 columns are undefined on census rows — a census is a
point measurement, not a sample). Read the distinct_senders_is_lower_bound
column together with distinct_senders: on rows where the bound
is true, the sender count is a lower bound from a cluster
sample, not an exact count. See
DATA-DICTIONARY.md for full
column semantics.
| chain | registry_address | method | window_start_block | window_end_block | window_start_utc | window_end_utc | registrations | distinct_senders | distinct_senders_is_lower_bound | top1_share | top10_share | top1_share_ci95_low | top1_share_ci95_high | top10_share_ci95_low | top10_share_ci95_high | hhi | sample_seed | n_resolved | notes |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ethereum | 0x8004A169FB4a3325136EB29fA0ceB6D2e539a432 | full_rescan | 24339871 | 25994481 | 2026-01-29T10:20:23Z | 2026-09-17T03:20:47Z | 50820 | 11352 | false | 0.049193 | 0.182487 | 50820 | |||||||
| base | 0x8004A169FB4a3325136EB29fA0ceB6D2e539a432 | full_rescan | 41663783 | 51415064 | 2026-02-03T11:08:33Z | 2026-09-17T04:31:15Z | 88500 | 29289 | false | 0.115831 | 0.272124 | 88500 | |||||||
| arbitrum | 0x8004A169FB4a3325136EB29fA0ceB6D2e539a432 | full_rescan | 428895443 | 505982474 | 2026-02-05T13:16:00Z | 2026-09-17T04:31:28Z | 1502 | 142 | false | 0.181758 | 0.391478 | 1502 | small_n_caveat | ||||||
| bnb_standard | 0x8004A169FB4a3325136EB29fA0ceB6D2e539a432 | full_raw_count_backfill | 79027268 | 122349961 | 352415 | 1289 | true | 0.096667 | 0.142 | 0.082726 | 0.112668 | 0.125249 | 0.16058 | 0.010379 | 659 | 1500 | early_era_only;cluster_sample_uniform_tuple;raw_count_is_whole_history | ||
| bnb_standard_late | 0x8004A169FB4a3325136EB29fA0ceB6D2e539a432 | cluster_sample_windowed | 104927268 | 122384655 | 1592 | 1117 | true | 0.0025 | 0.025 | 0.000851 | 0.007325 | 0.017567 | 0.035465 | 0.000963 | 660 | 1200 | late_era_only;cluster_sample_windowed | ||
| bnb_brc8004_fork | 0xfA09B3397fAC75424422C4D28b1729E3D4f659D7 | cluster_sample_windowed | 78864633 | 122424322 | 0 | 0 | true | 0 | 0 | 0 | 0 | 0 | 0 | 661 | 0 | fork_effectively_empty |
Downloads
| File | Type | Size | SHA-256 |
|---|---|---|---|
| dataset.csv | text/csv | 2,028 B | 678602836b8d10322db5da091017d4c5e42648d6e95fc3836ca0338a61f7870a |
| dataset.json | application/json | 5,478 B | 7639c6162935d3707afea2aeb358940c65b26c4191560640fb1da5b326f298bf |
| dataset.jsonld | application/ld+json | 7,935 B | 0d62be9c2087c5445442bf88e9c1ad653083a1b1b1c550a3c44ee89b0540319c |
| PROVENANCE.json | application/json | 13,824 B | 8513936fcf08676e4b47094071576ca77932e909ac57cca90f52feca96700323 |
| SHA256SUMS | text/plain | 320 B | |
Reading
| File | Purpose |
|---|---|
| DATA-DICTIONARY.md | Column-by-column reader's map (20 columns). |
| METHOD.md | Per-row window, RPC, seed, and reconciliation history. |
| GRADE.md | Self-grade against the seven claim traps. |
Related
Follow-up measurement: ERC-8004 Arbitrum One cohort census (public RPC) — 2026-10-07. First public-RPC-reproducible full-history scan of the Arbitrum IdentityRegistry + ReputationRegistry (1,604 agents from 1,000 owners; +6.79% registrations over this freeze's 1,502). Narrative companion: 1,604 ERC-8004 agents on Arbitrum, 108 with on-chain reputation — first public-RPC census.
If you need a stable columnar bulk-extract of our ERC-8004 datasets (chain-liveness and commerce-backed) with weekly Parquet delivery and a semver schema guarantee, see the dataset-licence schema & changelog (v1.0.0). Free CC BY 4.0 content on this page is unaffected.
Cite as
Onchain Agent Intel (2026). ERC-8004 IdentityRegistry census across Ethereum, Base, and Arbitrum One (current scope), with historical BNB Smart Chain rows retained for reproducibility (2026-09-17). CC BY 4.0. https://onchainagentintel.io/data/erc-8004-four-chain-census-2026-09
License
This dataset is published under Creative Commons Attribution 4.0 International (CC BY 4.0). Free to reuse, remix, and redistribute — including commercially — with attribution.