Skip to content

Benchmark

Horangi 4 subset: kmmlu

Higher scores are better.

Models measured
79
Scores recorded
83
Sources
1
IQ weight
1.95

Score distribution

0.776
Mean
0.82
Median
0.141
Std dev
0.61
10th percentile
0.91
90th percentile
0.21
Min
0.210.94
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.95 for chat models, measured over 79 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51MiMo V2 Promimo-v2-pro· Horangi Korean Leaderboard 4 data docs0.796
52Claude Haiku 4.5claude-haiku-4-5-20251001: high-effort· Horangi Korean Leaderboard 4 data docs0.77
53GLM 4.7 Flashglm-4.7-flash· Horangi Korean Leaderboard 4 data docs0.76
54GPT 4.1gpt-4.1· Horangi Korean Leaderboard 4 data docs0.76
55GPT-5 nanogpt-5-nano-2025-08-07· Horangi Korean Leaderboard 4 data docs0.76
56MiniMax M2.7minimax-m2.7· Horangi Korean Leaderboard 4 data docs0.75
57MiMo V2 Flashmimo-v2-flash· Horangi Korean Leaderboard 4 data docs0.747
58Mistral Medium 3.5mistral-medium-3-5· Horangi Korean Leaderboard 4 data docs0.74
59Qwen3.5 27Bqwen3.5-27b· Horangi Korean Leaderboard 4 data docs0.738
60MiniMax M2.5minimax-m2.5· Horangi Korean Leaderboard 4 data docs0.737
61GPT OSS 120Bgpt-oss-120b: high-effort· Horangi Korean Leaderboard 4 data docs0.73
62GPT-4.1 minigpt-4.1-mini· Horangi Korean Leaderboard 4 data docs0.72
63Qwen3.5 9Bqwen3.5-9b· Horangi Korean Leaderboard 4 data docs0.701
64Grok 4 Fastgrok-4-fast-non-reasoning· Horangi Korean Leaderboard 4 data docs0.7
65Nemotron 3 Nano 30B A3BNVIDIA-Nemotron-3-Nano-30B-A3B-BF16· Horangi Korean Leaderboard 4 data docs0.7
66Gemma 3 27B ITgemma-3-27b-it· Horangi Korean Leaderboard 4 data docs0.67
67GLM 4.5 Airglm-4.5-air· Horangi Korean Leaderboard 4 data docs0.67
68GPT 4ogpt-4o-2024-11-20· Horangi Korean Leaderboard 4 data docs0.67
69Grok 4.1 Fastgrok-4-1-fast-non-reasoning· Horangi Korean Leaderboard 4 data docs0.67
70Qwen3 14BQwen3-14B· Horangi Korean Leaderboard 4 data docs0.67
71Qwen3 8BQwen3-8B· Horangi Korean Leaderboard 4 data docs0.67
72Gemma 4 E4B-itgemma-4-E4B-it: enable-thinking· Horangi Korean Leaderboard 4 data docs0.65
73Mistral Smallmistral-small-2603· Horangi Korean Leaderboard 4 data docs0.62
74Qwen3 4B Instruct 2507Qwen3-4B-Instruct-2507· Horangi Korean Leaderboard 4 data docs0.62
75Gemma 4 E2B-itgemma-4-E2B-it: enable-thinking· Horangi Korean Leaderboard 4 data docs0.61
76GPT-4.1 nanogpt-4.1-nano· Horangi Korean Leaderboard 4 data docs0.6
77Gemma 3 12B ITgemma-3-12b-it· Horangi Korean Leaderboard 4 data docs0.57
78Qwen3 4BQwen3-4B· Horangi Korean Leaderboard 4 data docs0.57
79Qwen3 1.7BQwen3-1.7B· Horangi Korean Leaderboard 4 data docs0.56
80Gemma 3 4B ITgemma-3-4b-it· Horangi Korean Leaderboard 4 data docs0.47
81Qwen3 0.6BQwen3-0.6B· Horangi Korean Leaderboard 4 data docs0.38
82Gemma 3 1b itgemma-3-1b-it· Horangi Korean Leaderboard 4 data docs0.27
83Gemma 3 270M Itgemma-3-270m-it· Horangi Korean Leaderboard 4 data docs0.21

Showing 51–83 of 83 results.

Data credits: Horangi 4, Weights & Biases Korea; public publisher-selected numeric evaluation results