IKP - T5: leaderboard
IKP's fifth obscurity tier tests rare factual knowledge about Wikidata entities and researchers. Accuracy on all 181 cleaned T5 questions, answered without retrieval or supplied evidence.
Metric: Accuracy (%). Source: 01.me. Saturation forecast: Around February 2027. 138 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash (Preview) | 97.24 | #90 |
| 2 | Gemini 3 Flash (Preview) (Medium) | 97.24 | #90 (Gemini 3 Flash (Preview)) |
| 3 | Gemini 3.1 Pro (Preview) | 96.69 | #64 |
| 4 | GPT-5.5 | 95.03 | #33 |
| 5 | GPT-5.5 (Medium) | 93.37 | #33 (GPT-5.5) |
| 6 | Claude Fable 5 | 90.61 | #20 |
| 7 | Gemini 3.1 Flash Lite (Preview) | 86.74 | #201 |
| 8 | GPT-5.4 Pro (xHigh) | 85.64 | #38 (GPT-5.4 Pro) |
| 9 | MiMo-V2.5-Pro | 83.98 | #167 |
| 10 | GPT-5 (Medium) | 82.87 | #105 (GPT-5) |
| 11 | Gemini 2.5 Pro (Medium) | 82.87 | #159 (Gemini 2.5 Pro) |
| 12 | GPT-4.1 | 82.32 | #258 |
| 13 | O3 | 82.32 | #132 |
| 14 | Gemini 2.5 Pro (Preview) | 82.32 | #136 |
| 15 | GLM-5 | 81.77 | #149 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=ikp-t5 · How It Works · Data refreshed daily, snapshot 2026-11-05.