Model checks
Kimi K3, provider by provider. Model →
last audit 18 Nov · epoch 31 · evidence checked on-chain4 of 4
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
4
SAME
Failed
0
DIFF
Not tested
0
the audit could not run
By provider
Compared with moonshotai/kimi-k3 on OpenRouter · pinned to moonshotai · reasoning none
| Provider | Differed↑ | × noise | Fail line | Noise | Proof |
|---|---|---|---|---|---|
| Open BirdSAME | 8% | 2.0× | 13 under | 4.2% | ↗ |
| engySAME | 10% | 2.4× | 11 under | 4.2% | ↗ |
| BestSAME | 10% | 2.4× | 11 under | 4.2% | ↗ |
| RedPillSAME | 11% | 2.6× | 10 under | 4.2% | ↗ |
Evidence of some checks not found
29 older checks · 3 failed · not counting now
| Provider | Differed↑ | × noise | Fail line | Noise | Proof |
|---|---|---|---|---|---|
| Super SeederSAME 7 Nov · replaced | 4% | 0.5× | 30 under | 8.7% | ↗ |
| klinnSAME 7 Nov · expired | 5% | 0.5× | 29 under | 8.7% | ↗ |
| Open ForgeSAME 7 Nov · replaced | 5% | 0.5× | 29 under | 8.7% | ↗ |
| Venice.ai ProxySAME 7 Nov · replaced | 5% | 0.5× | 29 under | 8.7% | ↗ |
| Fire Ant 🔥🐜SAME 7 Nov · expired | 5% | 0.6× | 28 under | 8.7% | ↗ |
| antseed-crimson-swift-476bSAME 7 Nov · expired | 6% | 0.7× | 27 under | 8.7% | ↗ |
| bartly.eth64.deSAME 7 Nov · replaced | 6% | 0.7× | 27 under | 8.7% | ↗ |
| Prime SeedSAME 7 Nov · replaced | 7% | 0.8× | 26 under | 8.7% | ↗ |
| 💬 BanterSAME 7 Nov · expired | 7% | 0.8× | 26 under | 8.7% | ↗ |
| PhalaSAME 7 Nov · replaced | 7% | 0.8× | 26 under | 8.7% | ↗ |
| ZLKPro-ApiSAME 7 Nov · replaced | 7% | 0.8× | 20 under | 8.7% | ↗ |
| ▲ Apex AntSAME 7 Nov · replaced | 7% | 0.8× | 25 under | 8.7% | ↗ |
| dealerSAME 7 Nov · replaced | 8% | 0.9× | 18 under | 8.7% | ↗ |
| ApiportSAME 21 Oct · first round, before points | 16% | 1.7× | 4 under | 9.0% | ↗ |
| LLMartSAME 7 Nov · replaced | 16% | 1.8× | 12 under | 8.7% | ↗ |
| Edith AIDIFF 7 Nov · expired | 25% | 3.3× | +23 over | 7.7% | ↗ |
| Super Intelligence 21 Oct · no verdict this round | 26% | 2.9× | 1 under | 9.0% | ↗ |
| antseed-cosmic-gecko-c46dDIFF 7 Nov · expired | 28% | 3.2× | +6 over | 8.7% | ↗ |
| Vector Studios 7 Nov · 1 DIFF report, no verdict | 29% | 3.4× | +8 over | 8.7% | ↗ |
| Inspectant 7 Nov · 1 DIFF report, no verdict | 29% | 3.4× | +8 over | 8.7% | ↗ |
| Hana Gateway ✅DIFF 7 Nov · expired | 43% | 5.3× | +29 over | 8.0% | ↗ |
| Relay Works 7 Nov · replaced · not tested · provider unreachable | — | — | — | — | ↗ |
| attacompute.ai 7 Nov · expired · not tested · provider unreachable | — | — | — | — | ↗ |
| incharge-austin 7 Nov · replaced · not tested · provider unreachable | — | — | — | — | ↗ |
| Dark Signal 7 Nov · replaced · not tested · provider unreachable | — | — | — | — | ↗ |
| antseed-aggregator 7 Nov · expired · not tested · provider unreachable | — | — | — | — | ↗ |
| Vikey Seller 7 Nov · replaced · not tested · provider unreachable | — | — | — | — | ↗ |
| JOI SEED 21 Oct · not tested · provider unreachable | — | — | — | — | ↗ |
| NovaRoute AI 21 Oct · not tested · provider unreachable | — | — | — | — | ↗ |
Where providers miss
| Subject | Passed providers missed | Failed providers missed |
|---|---|---|
| Pop culture (box office and ratings) | 11% | — |
| Chinese geography (measurements) | 5% | — |
| Astronomy (periods and distances) | 10% | — |
| Biology (chromosome numbers) | 23% | — |
| Chemistry (melting points) | 14% | — |
| Chinese literature (publication years) | 0% | — |
| Programming (release years) | 6% | — |
| Chemistry (boiling points) | 9% | — |
| Medical (drug half-lives) | 13% | — |
| Chinese internet (platform dates) | 6% | — |
Audit rounds
| Round | Passed | Failed | Not tested |
|---|---|---|---|
| 18 Novepoch 31 | 4 | 0 | 0 |
| 7 Novepoch 30 | 18 | 3+2 waiting | 6 |
| 21 Octepoch 27 · before points | 3 | 2 | 11 |
Audited by approved auditors · evidence on IPFS, hash-checked on-chain.