Skip to content

Benchmark

KazBench Kazakh reading_comprehension (accuracy, dev)

Higher scores are better.

Models measured
6
Scores recorded
6
Sources
1
IQ weight
0.53

Score distribution

0.969
Mean
1
Median
0.045
Std dev
0.958
10th percentile
1
90th percentile
0.875
Min
0.8751
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.53 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 docs1
2Gemini 2.5 Flashdev, version: 0.1.0· KazBench Kazakh data docs1
3GPT-4o minidev, version: 0.1.0· KazBench Kazakh data docs1
4Qwen2.5 72B Instructdev, version: 0.1.0· KazBench Kazakh data docs0.979
5Llama 3.1 8B Instantdev, version: 0.1.0· KazBench Kazakh data docs0.958
6Llama 4 Scoutdev, version: 0.1.0· KazBench Kazakh data docs0.875

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.