Encyclo-K: leaderboard
Each question shows 8-10 textbook knowledge statements, some rewritten to be false, and the model picks which of 4-8 option combinations names exactly the statements asked for (the true ones or, in other questions, the false ones); the score is accuracy over 5,038 questions (3,286 English, 1,752 Chinese) across 11 disciplines. Unlike MMLU-style items that test one fact, every answer depends on judging several statements jointly.
Metric: Accuracy on all 5,038 questions (%). Source: encyclo-k.github.io. Saturation forecast: Around February 2027. 37 models tracked. Updated 17 Nov 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Qwen 3 235B A22B 2507 (Thinking) | 51.77 | #469 | |
| 2 | Qwen 3 235B A22B 2507 Instruct | 50.4 | #618 | |
| 3 | DeepSeek V3.1 Terminus | Thinking | 49.78 | #440 |
| 4 | Kimi K2 0905 | 44.66 | #564 | |
| 5 | Qwen 3 235B A22B | Thinking | 43.25 | #631 |
| 6 | QwQ-32B | 40.97 | #908 | |
| 7 | Qwen 3 30B A3B 2507 (Thinking) | 40.57 | #765 | |
| 8 | Qwen 3 32B | Thinking | 37.28 | #965 |
| 9 | Qwen 3 30B A3B 2507 Instruct | 36.72 | #1091 | |
| 10 | Qwen 3 30B A3B | Thinking | 31.56 | #1155 |
| 11 | DeepSeek V3 (0324) | 30.85 | #722 | |
| 12 | Qwen 3 14B | Reasoning | 30.79 | #1264 |
| 13 | Qwen 3 4B 2507 Instruct | 28.6 | #2215 | |
| 14 | Qwen 3 8B | Thinking | 26.78 | #1868 |
| 15 | Qwen 2.5 32B Instruct | 25.31 | #1257 |
No result here: #1 Claude Fable 5.2, #2 GPT-6.2 Astra, #3 Gemini 4 Carbon.
Interactive version: theaggregate.ai/benchmark?slug=encyclo-k · How It Works · Data refreshed daily, snapshot 2026-11-17.