Model checks
Claude Sonnet 5.5, provider by provider. Model →
last audit 18 Nov · epoch 31 · evidence checked on-chain12 of 15
providers answered as the model
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
12
SAME
Failed
3
DIFF
Not tested
3
the audit could not run
By provider
Compared with anthropic/claude-sonnet-5.5 on OpenRouter · pinned to anthropic · reasoning minimal
| Provider | Differed↑ | × noise | Fail line | Noise | Proof |
|---|---|---|---|---|---|
| bartly.eth64.deSAME auditors split: 1 DIFF, 1 SAME | 2% | 0.7× | 17 under | 2.5% | ↗ |
| ▲ Apex AntSAME | 4% | 0.8× | 20 under | 4.6% | ↗ |
| ZLKPro-ApiSAME | 5% | 1.2× | 18 under | 4.6% | ↗ |
| Super SeederSAME | 6% | 1.3× | 17 under | 4.6% | ↗ |
| Unified Rack · DeepSeek 240 tok/sSAME | 6% | 1.3× | 17 under | 4.6% | ↗ |
| Lab StudiosSAME | 6% | 1.3× | 17 under | 4.6% | ↗ |
| Relay WorksSAME | 6% | 1.3× | 17 under | 4.6% | ↗ |
| GesundAISAME | 6% | 1.3× | 17 under | 4.6% | ↗ |
| STEYGSAME | 7% | 1.5× | 16 under | 4.6% | ↗ |
| klinnSAME | 7% | 1.5× | 16 under | 4.6% | ↗ |
| AI Inference GatewaySAME | 8% | 1.7× | 15 under | 4.6% | ↗ |
| LLMartSAME | 11% | 2.3× | 11 under | 4.6% | ↗ |
| BestDIFF | 27% | 11× | +13 over | 2.5% | ↗ |
| Vikey SellerDIFF | 36% | 14× | +24 over | 2.5% | ↗ |
| TokenSaleDIFF | 57% | 23× | +50 over | 2.5% | ↗ |
| attacompute.ai not tested · timed out | — | — | — | — | ↗ |
| antseed-aggregator not tested · provider unreachable | — | — | — | — | ↗ |
| Venice.ai Proxy not tested · provider unreachable | — | — | — | — | ↗ |
9 older checks · 3 failed · not counting now
| Provider | Differed↑ | × noise | Fail line | Noise | Proof |
|---|---|---|---|---|---|
| 💬 BanterSAME 7 Nov · expired | 1% | 0.5× | 16 under | 2.0% | ↗ |
| antseed-crimson-swift-476bSAME 7 Nov · expired | 4% | 2.0× | 13 under | 2.0% | ↗ |
| Fire Ant 🔥🐜SAME 7 Nov · expired | 4% | 2.0× | 13 under | 2.0% | ↗ |
| ApiportSAME 21 Oct · first round, before points | 11% | 1.1× | 12 under | 9.8% | ↗ |
| Hana Gateway ✅DIFF 7 Nov · expired | 21% | 6.3× | +4 over | 3.3% | ↗ |
| Inspectant 7 Nov · expired · 1 DIFF, the 2nd audit could not run | 21% | 11× | +4 over | 2.0% | ↗ |
| antseed-cosmic-gecko-c46dDIFF 7 Nov · expired | 25% | 7.5× | +9 over | 3.3% | ↗ |
| dealerDIFF 7 Nov · expired | 30% | 9.0× | +15 over | 3.3% | ↗ |
| RouteIT 4 Nov · expired · not tested · provider unreachable | — | — | — | — | ↗ |
Where providers miss
| Subject | Passed providers missed | Failed providers missed |
|---|---|---|
| Biology (chromosome numbers) | 7% | 46% |
| Medical (drug half-lives) | 12% | 38% |
| Programming (release years) | 6% | 35% |
| Astronomy (periods and distances) | 2% | 32% |
| Chinese geography (measurements) | 9% | 32% |
| Chemistry (melting points) | 11% | 38% |
| Mathematics (special values) | 6% | 36% |
| Physics (material properties) | 11% | 21% |
| Pop culture (box office and ratings) | 5% | 36% |
| Chinese literature (publication years) | 0% | 39% |
Audit rounds
| Round | Passed | Failed | Not tested |
|---|---|---|---|
| 17 Novepoch 31 | 12 | 3 | 3 |
| 7 Novepoch 30 | 12 | 5 | 5 |
| 4 Novepoch 29 | 12 | 6+1 waiting | 3 |
| 21 Octepoch 27 · before points | 8 | 3 | 0 |
Audited by approved auditors · evidence on IPFS, hash-checked on-chain.