Encyclo-K - Engineering: leaderboard

Encyclo-K's multi-statement selection questions restricted to Engineering: accuracy over the 881 Engineering questions of the 5,038-question standard set.

Metric: Accuracy on the 881 Engineering 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)54.71#469
2Qwen 3 235B A22B 2507 Instruct51.42#618
3DeepSeek V3.1 TerminusThinking49.6#440
4QwQ-32B44.49#908
5Qwen 3 235B A22BThinking44.38#631
6Kimi K2 090543.47#564
7Qwen 3 30B A3B 2507 (Thinking)43.25#765
8Qwen 3 32BThinking40.41#965
9Qwen 3 30B A3B 2507 Instruct38.14#1091
10Qwen 3 14BReasoning34.17#1264
11Qwen 3 30B A3BThinking33.48#1155
12DeepSeek V3 (0324)31.78#722
13Qwen 3 4B 2507 Instruct31.44#2215
14Qwen 3 8BThinking29.06#1868
15Qwen 2.5 32B Instruct27.81#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-engineering · How It Works · Data refreshed daily, snapshot 2026-11-17.