Skip to content

Benchmark

Horangi 4 subset: ko_mtbench

People compare model answers and choose which they prefer. Higher scores are better.

Models measured
79
Scores recorded
83
Sources
1
IQ weight
1.76

Score distribution

0.085
Mean
0.087
Median
0.009
Std dev
0.081
10th percentile
0.089
90th percentile
0.0234
Min
0.02340.0901
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.76 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
51Qwen3.6 Flashqwen3.6-flash· Horangi Korean Leaderboard 4 data docs0.0861
52Grok 4.5grok-4.5· Horangi Korean Leaderboard 4 data docs0.086
53MiniMax M2.7minimax-m2.7· Horangi Korean Leaderboard 4 data docs0.0858
54GPT-5.6 Solgpt-5.6-sol: max-effort· Horangi Korean Leaderboard 4 data docs0.0856
55Qwen3.5 35B A3Bqwen3.5-35b-a3b· Horangi Korean Leaderboard 4 data docs0.0855
56GPT-4.1 minigpt-4.1-mini· Horangi Korean Leaderboard 4 data docs0.0853
57MiniMax M2.5minimax-m2.5· Horangi Korean Leaderboard 4 data docs0.0852
58GPT-5 nanogpt-5-nano-2025-08-07· Horangi Korean Leaderboard 4 data docs0.085
59Nemotron 3 Super 120B A12BNVIDIA-Nemotron-3-Super-120B-A12B-FP8· Horangi Korean Leaderboard 4 data docs0.0849
60Qwen3 14BQwen3-14B· Horangi Korean Leaderboard 4 data docs0.0849
61Gemma 3 27B ITgemma-3-27b-it· Horangi Korean Leaderboard 4 data docs0.0846
62Qwen3 4B Instruct 2507Qwen3-4B-Instruct-2507· Horangi Korean Leaderboard 4 data docs0.0845
63Grok 4 Fastgrok-4-fast-non-reasoning· Horangi Korean Leaderboard 4 data docs0.0844
64Qwen3 32BQwen3-32B· Horangi Korean Leaderboard 4 data docs0.0844
65Grok 4 Fastgrok-4-fast-reasoning· Horangi Korean Leaderboard 4 data docs0.0843
66Nemotron 3 Nano 30B A3BNVIDIA-Nemotron-3-Nano-30B-A3B-BF16· Horangi Korean Leaderboard 4 data docs0.0843
67Grok 4.1 Fastgrok-4-1-fast-non-reasoning· Horangi Korean Leaderboard 4 data docs0.084
68Qwen3 8BQwen3-8B· Horangi Korean Leaderboard 4 data docs0.084
69MiMo V2 Flashmimo-v2-flash· Horangi Korean Leaderboard 4 data docs0.084
70Qwen3.5 27Bqwen3.5-27b· Horangi Korean Leaderboard 4 data docs0.0838
71Mistral Smallmistral-small-2603· Horangi Korean Leaderboard 4 data docs0.0833
72GPT-4.1 nanogpt-4.1-nano· Horangi Korean Leaderboard 4 data docs0.0825
73GLM 4.7 Flashglm-4.7-flash· Horangi Korean Leaderboard 4 data docs0.0823
74Gemma 3 12B ITgemma-3-12b-it· Horangi Korean Leaderboard 4 data docs0.0811
75Qwen3.5 9Bqwen3.5-9b· Horangi Korean Leaderboard 4 data docs0.081
76Qwen3 4BQwen3-4B· Horangi Korean Leaderboard 4 data docs0.0798
77Mistral Medium 3.5mistral-medium-3-5· Horangi Korean Leaderboard 4 data docs0.0788
78Gemma 3 4B ITgemma-3-4b-it· Horangi Korean Leaderboard 4 data docs0.078
79Qwen3 1.7BQwen3-1.7B· Horangi Korean Leaderboard 4 data docs0.0748
80Grok 4.1 Fastgrok-4-1-fast-reasoning· Horangi Korean Leaderboard 4 data docs0.0685
81Gemma 3 1b itgemma-3-1b-it· Horangi Korean Leaderboard 4 data docs0.0561
82Qwen3 0.6BQwen3-0.6B· Horangi Korean Leaderboard 4 data docs0.0485
83Gemma 3 270M Itgemma-3-270m-it· Horangi Korean Leaderboard 4 data docs0.0234

Showing 51–83 of 83 results.

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