Open CoT - LSAT Reading Comprehension — leaderboard

Metric: CoT Gain (%). Source: huggingface.co. 132 models tracked.

Top models

#ModelScore
1Llama 3.1 8B Instruct27.51
2Llama 3 8B Instruct25.65
3aya-expanse-32B24.54
4internlm2-chat-20B24.16
5Llama 3.1 70B Instruct23.42
6Daredevil-8B-abliterated23.42
7Ministral-8B-Instruct-241023.05
8DeepSeek R1 Distill Llama 8B22.68
9SOLAR-10.7B-Instruct-v1.022.3
10internlm2-chat-7B21.56
11DeepSeek R1 Distill Qwen 14B21.19
12DBRX21.19
13Llama 3 70B Instruct20.82
14DeepSeek-R1-Distill-Qwen-7B20.82
15Yi 1.5 34B Chat20.45

Interactive version: theaggregate.ai/benchmark?slug=open-cot-lsat-reading-comprehension · How the rankings work · Data refreshed daily, snapshot 2026-07-22.