IKP - T7: leaderboard
IKP's hardest obscurity tier tests extreme long-tail factual recall about entities and researchers. Accuracy on all 172 cleaned T7 questions, answered without retrieval or supplied evidence; missing evaluations are not counted as wrong answers.
Metric: Accuracy (%). Source: 01.me. Saturation forecast: Around 2031. 137 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash (Preview) (Medium) | 28.49 | #90 (Gemini 3 Flash (Preview)) |
| 2 | Gemini 3 Flash (Preview) | 23.84 | #90 |
| 3 | Gemini 2.5 Pro (Preview) | 22.09 | #136 |
| 4 | Gemini 2.5 Pro (Medium) | 21.51 | #159 (Gemini 2.5 Pro) |
| 5 | GPT-4.1 | 20.93 | #258 |
| 6 | GPT-5.5 (Medium) | 20.93 | #33 (GPT-5.5) |
| 7 | GPT-5.5 | 20.35 | #33 |
| 8 | Claude Fable 5 | 20.35 | #20 |
| 9 | GLM-5 | 19.19 | #149 |
| 10 | GPT-5 Pro | 18.6 | #89 |
| 11 | O3 | 18.02 | #132 |
| 12 | GPT-5.4 Pro (xHigh) | 18.02 | #38 (GPT-5.4 Pro) |
| 13 | GPT-5 | 16.86 | #105 |
| 14 | GPT-4.1 Mini | 16.86 | #363 |
| 15 | Gemini 2.0 Flash (001) | 16.86 | #365 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=ikp-t7 · How It Works · Data refreshed daily, snapshot 2026-11-05.