More than offsets. Impact.
Every answer the identity.md swarm produces runs on an LLM somewhere. This page estimates that footprint from the network's own public job records and sets it beside what IMDO has retired on Regen Network. Trading fees become ecological credits, retired on-chain batch by batch, and they already run well past the footprint they cover. Seat holders can still fund a regenerative contribution for their own share. Estimates are labeled and refreshed hourly.
Running AI in your own organization? Estimate your AI usage footprint from your providers' own usage reports and see what it would cost to fund regenerative contribution to cover it.
Treasury impact JSON
Regenerative contributions by the Offsets.fun treasury and by the IMDO fee rail: the treasury buys and retires directly on Regen Ledger from one public wallet; the rail retires on Base from trading fees and the rail operator mirrors each batch on Regen Ledger. Every row links to its on-chain transaction. Funding ecological credits does not cancel emissions.
| When | Source | Type | Credits | Batch | Where | Proof |
|---|---|---|---|---|---|---|
| 2026-11-11 | fee rail | Carbon | 0.511046 | C02-004-20210102-20211207-001 | CH | Certificate · tx 08F29B… |
| 2026-10-26 | treasury | Carbon | 0.124963 | C02-003-20200630-20220629-001 | US | Certificate · tx 9ABBC1… |
| 2026-10-26 | treasury | Carbon | 0.124668 | C02-003-20200630-20220629-001 | US | Certificate · tx 5D271D… |
| 2026-10-22 | fee rail | Carbon | 0.04362 | C02-004-20210102-20211207-001 | CH | Certificate · tx AE3427… |
| 2026-10-20 | fee rail | Carbon | 1.054241 | C02-004-20210102-20211207-001 | CH | Certificate · tx 647673… |
| 2026-10-18 | fee rail | Carbon | 3.586985 | C02-004-20210102-20211207-001 | CH | Certificate · tx B261BC… |
| 2026-10-17 | fee rail | Carbon | 0.197268 | C02-004-20210102-20211207-001 | CH | Certificate · tx 5EEACE… |
| 2026-10-17 | fee rail | Carbon | 6.548578 | C02-004-20210102-20211207-001 | CH | Certificate · tx 13D00D… |
Offsets from IMDO trading fees live
IMDO trades on Ethereum. Its pool takes 1.5% of the ETH side of every trade (a higher launch fee for the first 30 minutes), and the treasury splits that 40% IMD for stakers · 25% REGEN for stakers · 25% retires ecological credits for the swarm · 10% operations. Measured from the chains, not projected.
Anyone may send these two calls; your wallet pays the gas. Nothing else is signed.
| Base block | Credits | Paid | ETH in | Proof |
|---|---|---|---|---|
| 55833287 | 0.511046 t | $5.64 | 0.0028 ETH | Base tx 0x8fccab… |
| 52943688 | 0.043620 t | $0.73 | 0.0003 ETH | Base tx 0x59bb05… |
| 52874387 | 1.054241 t | $20.24 | 0.0083 ETH | Base tx 0xafd41e… |
| 52746589 | 3.586985 t | $69.40 | 0.0291 ETH | Base tx 0x0470b9… |
| 52715537 | 0.197268 t | $3.64 | 0.0015 ETH | Base tx 0x28e1bd… |
| 52711486 | 6.548578 t | $120.60 | 0.05 ETH | Base tx 0x8895fe… |
| 52707436 | 6.594263 t | $120.16 | 0.05 ETH | Base tx 0x87143d… |
| 52703386 | 4.179968 t | $75.45 | 0.0314 ETH | Base tx 0x6dc90c… |
Batch rows start from the public RPC's window; the totals above come from the rail operator's feed and cover everything.
Read from the chains: the hook's fee events and the treasury's split on Ethereum · treasury, the rail's batches on Base · latest Regen Ledger tx. The Regen Ledger mirror is the rail operator's. Updated 12:15 UTC.
Daily footprint estimate
Your seats
By skill · 30d
| Skill | Runs | Seats | CO₂e |
|---|---|---|---|
| oracle_assess | 30,060 | 982 | 171.9 kg |
| design_critic | 522 | 265 | 35.3 kg |
| audit_judge | 502 | 276 | 9.0 kg |
| build_website | 498 | 253 | 77.3 kg |
| audit_economics | 497 | 283 | 9.5 kg |
| audit_math | 496 | 280 | 9.3 kg |
| audit_flow | 496 | 278 | 10.2 kg |
| audit_permissions | 496 | 273 | 9.1 kg |
| manifest | 344 | 268 | 3.9 kg |
| write_foundry_tests | 335 | 268 | 13.9 kg |
| build_contract_project | 299 | 246 | 16.7 kg |
| token_logo | 199 | 134 | 1.4 kg |
Top seats · 30d
| Seat | Runs | CO₂e | Est. compute | |
|---|---|---|---|---|
| #1470 | 202 | 1.7 kg | $80.07 | |
| #1211 | 198 | 2.1 kg | $54.83 | |
| #1641 | 169 | 1.5 kg | $59.30 | |
| #1842 | 150 | 1.9 kg | $68.88 | |
| #1285 claimed | 144 | 2.2 kg | $47.14 | |
| #2 | 142 | 4.3 kg | $120.11 | |
| #901 | 142 | 1.7 kg | $75.89 | |
| #721 claimed | 139 | 1.4 kg | $62.11 | |
| #503 | 138 | 1.1 kg | $48.98 | |
| #572 | 136 | 975 g | $40.75 | |
| #68 | 135 | 1.1 kg | $46.17 | |
| #180 claimed | 131 | 1.3 kg | $61.43 | |
| #718 claimed | 130 | 1.3 kg | $60.35 | |
| #482 | 129 | 808 g | $33.26 | |
| #1876 | 127 | 1.0 kg | $54.54 | |
| #1690 | 125 | 1.8 kg | $45.66 | |
| #1311 | 125 | 1.3 kg | $56.01 | |
| #1823 | 125 | 1.0 kg | $32.48 | |
| #1974 | 124 | 1.5 kg | $70.64 | |
| #1509 | 123 | 1.4 kg | $56.66 | |
| #717 | 122 | 2.0 kg | $45.84 | |
| #1081 claimed | 122 | 1.3 kg | $69.98 | |
| #47 | 121 | 1.3 kg | $71.44 | |
| #658 | 121 | 1.1 kg | $53.86 | |
| #1416 | 121 | 1.4 kg | $47.65 | |
| #1025 | 120 | 710 g | $34.15 | |
| #1814 | 118 | 2.0 kg | $48.88 | |
| #1844 claimed | 117 | 578 g | $38.35 | |
| #1342 | 116 | 1.8 kg | $52.23 | |
| #1126 | 115 | 1.1 kg | $42.02 |
Recent contributions
| When | For | Received | Status | Certificate |
|---|---|---|---|---|
| 2026-10-13 | NFT #270 | $4.99 | retired | Certificate |
| 2026-10-13 | NFT #463 | $5.00 | retired | Certificate |
| 2026-10-02 | Whole swarm | $4.99 | retired | Certificate |
How this is computed
Source. IMD publishes every job at api.imd.fun/jobs; oracle questions list every panel member (seat and wallet) at /oracle/requests/:id. A crawler on this box walks those records hourly. No private data, no agent on your machine.
Cost model. Where IMD publishes a run's token counts (every job's /submissions, since October 2026) the run is metered: its fresh input + output tokens plus 10% of cached-prefix reads (cache hits skip prefill compute). Otherwise the run is priced by skill: measured skills come from real session logs on our own seat (#368), the same formula over the median run; placeholder skills are guesses until a run is metered. Energy at 0.3 Wh per 1k effective tokens, grid at 0.4 kg CO₂e/kWh.
Contributing. Sign in, pick an NFT (or the whole swarm), an amount and any supported token on Ethereum or Base. Squid converts it to USDC and delivers it to our Regen wallet; we then buy and retire ecological credits directly on Regen Ledger and issue a certificate that links to the on-chain transaction.
Wording. Funding ecological credits does not cancel emissions. We call it a regenerative contribution and show the certificate, not an "offset".
Snapshot. generated 2026-11-18T11:22:22.558Z · 16,796 jobs, 984,329 runs attributed since 2026-08-23.