Model checks
Gemini 4 Argon, seller by seller. Model →
last audit 7 Nov · epoch 30 · evidence checked on-chain12 of 13
sellers answered as the model
How this works (Antseed's announcement ↗)
Approved auditors check that each seller 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 seller 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 seller 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 seller 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
1
DIFF
Not tested
2
the audit could not run
By seller
Compared with google/gemini-4-argon on OpenRouter · pinned to google-ai-studio · reasoning minimal
| Seller | Differed↑ | × noise | Fail line | Noise | Proof |
|---|---|---|---|---|---|
| Super SeederSAME | 0% | 0.0× | 14 under | 1.0% | ↗ |
| InspectantSAME | 1% | 1.0× | 13 under | 1.0% | ↗ |
| Fire Ant 🔥🐜SAME | 1% | 1.0× | 13 under | 1.0% | ↗ |
| ▲ Apex AntSAME | 1% | 1.0× | 13 under | 1.0% | ↗ |
| 💬 BanterSAME | 1% | 1.0× | 13 under | 1.0% | ↗ |
| Hana Gateway ✅SAME | 1% | 1.0× | 13 under | 1.0% | ↗ |
| Unified Rack · DeepSeek 240 tok/sSAME | 1% | 1.0× | 13 under | 1.0% | ↗ |
| LLMartSAME | 1% | 1.0× | 13 under | 1.0% | ↗ |
| attacompute.aiSAME | 1% | 1.1× | 13 under | 1.0% | ↗ |
| ZLKPro-ApiSAME | 2% | 2.0× | 12 under | 1.0% | ↗ |
| OpenCheapSAME | 3% | 3.0× | 11 under | 1.0% | ↗ |
| BestSAME | 8% | 8.0× | 6 under | 1.0% | ↗ |
| Edith AIDIFF | 25% | 25× | +11 over | 1.0% | ↗ |
| zro not tested · seller unreachable | — | — | — | — | ↗ |
| antseed-aggregator not tested · seller unreachable | — | — | — | — | ↗ |
Where sellers miss
| Subject | Passed sellers missed | Failed sellers missed |
|---|---|---|
| Biology (chromosome numbers) | 0% | 60% |
| Chinese internet (platform dates) | 1% | 14% |
| Pop culture (box office and ratings) | 4% | 11% |
| Programming (release years) | 3% | 47% |
| Chinese history (specific dates) | 0% | 38% |
| Physics (material properties) | 2% | 11% |
| Astronomy (periods and distances) | 3% | 6% |
| Chinese geography (measurements) | 0% | 6% |
| Medical (drug half-lives) | 0% | 13% |
| Chinese literature (publication years) | 8% | 18% |
Audit rounds
| Round | Passed | Failed | Not tested |
|---|---|---|---|
| 7 Novepoch 30 | 12 | 1 | 2 |
| 4 Novepoch 29 | 9 | 2 | 1 |
Audited by approved auditors · evidence on IPFS, hash-checked on-chain.