IKP - T4: leaderboard
IKP's fourth obscurity tier tests specialized factual recall, including obscure entities and researchers. Accuracy on all 188 cleaned T4 questions; no retrieval, supplied evidence or hallucination penalty.
Metric: Accuracy (%). Source: 01.me. Saturation forecast: Estimated already saturated. 196 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 98.4 | #64 |
| 2 | Claude Opus 4.6 (High) | 96.81 | #67 (Claude Opus 4.6) |
| 3 | GPT-5.5 | 96.28 | #33 |
| 4 | GPT-5.5 (Medium) | 96.28 | #33 (GPT-5.5) |
| 5 | Gemini 3 Flash (Preview) (Medium) | 96.28 | #90 (Gemini 3 Flash (Preview)) |
| 6 | Claude Opus 4.6 | 95.74 | #67 |
| 7 | Gemini 3 Flash (Preview) | 95.74 | #90 |
| 8 | Gemini 2.5 Pro (Preview) | 95.21 | #136 |
| 9 | MiMo-V2.5-Pro | 94.68 | #167 |
| 10 | Claude Fable 5 | 94.15 | #20 |
| 11 | GPT-5 Pro | 94.15 | #89 |
| 12 | GPT-5.4 Pro (xHigh) | 94.15 | #38 (GPT-5.4 Pro) |
| 13 | Gemini 2.5 Pro (Medium) | 94.15 | #159 (Gemini 2.5 Pro) |
| 14 | Grok 4 | 93.62 | #189 |
| 15 | Grok 3 | 93.62 | #313 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=ikp-t4 · How It Works · Data refreshed daily, snapshot 2026-11-05.