Skip to content

Benchmark

KazBench Kazakh Overall (dev)

Higher scores are better.

Models measured
6
Scores recorded
6
Sources
1
IQ weight
0.92

Score distribution

81.8
Mean
87.5
Median
9.2
Std dev
74.5
10th percentile
91.5
90th percentile
64.9
Min

This benchmark has a weight of 0.92 for chat models, measured over 6 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Claude Haiku 3.5dev, version: 0.1.0· KazBench Kazakh data docs91.5
2GPT-4o minidev, version: 0.1.0· KazBench Kazakh data docs88.3
3Llama 4 Scoutdev, version: 0.1.0· KazBench Kazakh data docs87.5
4Gemini 2.5 Flashdev, version: 0.1.0· KazBench Kazakh data docs84.1
5Qwen2.5 72B Instructdev, version: 0.1.0· KazBench Kazakh data docs74.5
6Llama 3.1 8B Instantdev, version: 0.1.0· KazBench Kazakh data docs64.9

Data credits: KazBench by Yersultan Akhmer. Provisional DEV numeric results; code MIT and task-data CC BY 4.0 do not establish a separate score licence.