Encyclo-K - Medicine: leaderboard

Encyclo-K's multi-statement selection questions restricted to Medicine: accuracy over the 736 Medicine questions of the 5,038-question standard set.

Metric: Accuracy on the 736 Medicine questions (%). Source: encyclo-k.github.io. Saturation forecast: Around February 2027. 37 models tracked. Updated 17 Nov 2026.

Top models

#ModelSettingScoreOverall rank
1Qwen 3 235B A22B 2507 (Thinking)54.48#469
2Qwen 3 235B A22B 2507 Instruct52.17#618
3DeepSeek V3.1 TerminusThinking52.17#440
4Qwen 3 235B A22BThinking48.91#631
5Kimi K2 090545.79#564
6QwQ-32B44.84#908
7Qwen 3 30B A3B 2507 (Thinking)44.02#765
8Qwen 3 32BThinking40.62#965
9Qwen 3 30B A3B 2507 Instruct37.5#1091
10Qwen 3 30B A3BThinking33.56#1155
11DeepSeek V3 (0324)32.34#722
12Qwen 3 14BReasoning32.34#1264
13Qwen 3 4B 2507 Instruct29.08#2215
14Qwen 3 8BThinking27.31#1868
15Qwen 2.5 72B Instruct25.41#1119

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-medicine · How It Works · Data refreshed daily, snapshot 2026-11-17.