🐜AntSeedStats
HomeNetworkModelsProvidersStakingFree AI$ANTS

Explore

OverviewRevenueProvidersBuyersModelsDistributionStakingRewardsDIEM$ANTSChannelsTransactionsLabs

Project

AboutFAQDevelopersFor agentsChangelogStatus

Antseed

WebsiteDownload Antseed DesktopDocsGitHubX (@antseedai)Telegram
🐜AntSeedStats
The metrics and intelligence layer of the Antseed ecosystem.
Built by @gekko_eth
AntSeedStats is an independent, community-built dashboard, not officially affiliated with the Antseed team. All data is sourced on-chain and from public endpoints.

Model checks

TokenSale, as the auditors found it. Provider →

last audit 18 Nov · evidence checked on-chain
×0.64

on its $ANTS points now

How this works (Antseed's announcement ↗)

Approved auditors check that each provider serves the model it sells.

  • 🧪They buy from it like any buyer: ~100 obscure number questions, compared with the real model.
  • 📏The real model misses some too: that is the noise. A provider fails only if it misses clearly more, and two auditors must agree.
  • 💰Each pass +2% on $ANTS points, up to +50%. Fails cost more: 1 −4%, 2 −16%, 3 −36%. Checked again every week.
How this works (Antseed's announcement ↗)

Approved auditors check that each provider serves the model it sells.

  • 🧪They buy from it like any buyer: ~100 obscure number questions, compared with the real model.
  • 📏The real model misses some too: that is the noise. A provider fails only if it misses clearly more, and two auditors must agree.
  • 💰Each pass +2% on $ANTS points, up to +50%. Fails cost more: 1 −4%, 2 −16%, 3 −36%. Checked again every week.
Passed
0
SAME · counting now
Failed
3
DIFF
Last audit
18 Nov
epoch 31 · counts to 32
By model
ModelDiffered↑× noiseFail lineNoiseProof
Claude Fable 5.1DIFF43%8.7×+34 over5.0%↗
Claude Opus 5.5DIFF49%54×+35 over0.9%↗
Claude Sonnet 5.5DIFF57%23×+50 over2.5%↗
Why it failed
Claude Sonnet 5.5DIFF

31 of 130 differed (fails from 25) · 69 of 120 differed (fails from 19) · the real model differs from itself on 3%

Where it missed
Chinese literature (publication years)11/19
Mathematics (special values)11/20
Medical (drug half-lives)15/29
Biology (chromosome numbers)13/28
For example
Elimination half-life of lumasiran (hours)real 5 · provider 60
Ramanujan tau function value tau(10)real -115,920 · provider 534,612
Sidereal orbital period of Siarnaq around Saturn in daysreal 896 · provider 3,408

✓Reasoning switched off with reasoning effort none, as the real model was asked.

audited 18 Nov, 18:12 UTC · epoch 31
test power 93% · median answer 5.2 s · 2 audits
Claude Opus 5.5DIFF

41 of 120 differed (fails from 17) · 49 of 100 differed (fails from 14) · the real model differs from itself on 1%

Where it missed
Biology (chromosome numbers)33/54
Astronomy (periods and distances)12/22
Physics (material properties)3/7
Chinese internet (platform dates)11/27
For example
siponimod elimination half life hoursreal 30 · provider 240
Southern Crab Nebula Henize 2 catalog number (Hen 2 number)real 104 · provider 302
salvinia cucullata diploid chromosome number 2n countreal 18 · provider 48

✓Reasoning switched off with reasoning effort none, as the real model was asked.

audited 18 Nov, 18:13 UTC · epoch 31
test power 96% · median answer 7.0 s · 2 audits
Claude Fable 5.1DIFF

46 of 110 differed (fails from 21) · 60 of 138 differed (fails from 26) · the real model differs from itself on 5%

Where it missed
Mathematics (special values)2/3
Physics (material properties)10/17
Pop culture (box office and ratings)5/9
Medical (drug half-lives)25/48
For example
M87 jet projected length in kiloparsecsreal 1.5 · provider 5
Qaidam Basin area (square kilometers)real 120,000 · provider 257,000
trofinetide mean elimination half life hoursreal 1.5 · provider 3

✓Reasoning switched off with reasoning effort none, as the real model was asked.

audited 18 Nov, 18:13 UTC · epoch 31
test power 95% · median answer 7.6 s · 2 audits
3 older checks · 2 failed · not counting now
ModelDiffered↑× noiseFail lineNoiseProof
GPT-6.1 SolSAME 4 Nov · expired2%1.2×9 under1.7%↗
GPT-6 AstraDIFF 7 Nov · expired42%23×+29 over1.8%↗
GLM 5.3DIFF 7 Nov · expired55%4.7×+32 over11.5%↗
GLM 5.3DIFFno longer counts

58 of 128 differed (fails from 33) · 71 of 130 differed (fails from 39) · the real model differs from itself on 12%

Where it missed
Chemistry (melting points)13/16
Medical (drug half-lives)13/19
Astronomy (periods and distances)9/16
Chinese literature (publication years)9/16
For example
fifth bernoulli number b5 numeratorreal 0 · provider -1
cast 128 block size bitsreal 64 · provider 128
black hole to galaxy bulge mass ratio of ngc 1277 in percentreal 14 · provider 0.04

✓Reasoning switched off with reasoning effort none, as the real model was asked.

audited 7 Nov, 22:26 UTC · epoch 30
test power 92% · median answer 3.8 s · 2 audits
GPT-6 AstraDIFFno longer counts

46 of 110 differed (fails from 17) · 38 of 110 differed (fails from 25) · the real model differs from itself on 2%

Where it missed
Physics (material properties)8/11
Pop culture (box office and ratings)9/13
Medical (drug half-lives)13/21
Programming (release years)17/36
For example
SCHOTT N-BK7 Sellmeier C2 squared-resonance-wavelength coefficient, catalog dispersion formula (10^-12 m^2)real 2.002e-14 · provider 0.02002
SCHOTT N-SF11 catalog Sellmeier C1 (10^-12 m^2)real 1.319e-14 · provider 0.01319
Fused silica Sellmeier third resonance wavelength, Malitson 1965 fit at 293.15 K (10^-6 m)real 0.000009896 · provider 9.896

✓Reasoning switched off with reasoning effort none, as the real model was asked.

audited 7 Nov, 22:13 UTC · epoch 30
test power 95% · median answer 14.4 s · 2 audits
Audit rounds
RoundPassedFailedNot testedMultiplier
18 Novepoch 31030×0
7 Novepoch 30050×0
4 Novepoch 29240×0.40
21 Octepoch 27 · before points002—

Audited by approved auditors · evidence on IPFS, hash-checked on-chain.