Skip to content

Benchmark

IdrockBench Uzbek reasoning (LiveBench partial credit, task v2.0)

Scores are percentages. Higher is better.

Models measured
12
Scores recorded
12
Sources
1
IQ weight
0.77

Score distribution

41%
Mean
42.4%
Median
28.4%
Std dev
7.9%
10th percentile
73.8%
90th percentile
1.4%
Min
1.4%87.7%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.77 for chat models, measured over 12 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Qwen3.6 27BQ4_K_M, reasoning: off, organiser-run· IdrockBench Uzbek leaderboard data docs87.7%
2Qwen3.5 27BQ4_K_M, reasoning: off, organiser-run· IdrockBench Uzbek leaderboard data docs73.8%
3Qwen 3.8 27BQ4_K_M, reasoning: off, organiser-run· IdrockBench Uzbek leaderboard data docs73.6%
4Gemma 4 31B ITQ4_K_M, reasoning: off, organiser-run· IdrockBench Uzbek leaderboard data docs67.8%
5Gemma 4 26BQ4_K_M, reasoning: off, organiser-run· IdrockBench Uzbek leaderboard data docs52%
6Qwen3.5 9BQ4_K_M, reasoning: off, organiser-run· IdrockBench Uzbek leaderboard data docs42.4%
7gemma 4 E4BQ4_K_M, reasoning: off, organiser-run· IdrockBench Uzbek leaderboard data docs33.9%
8Qwen3.5 4BQ4_K_M, reasoning: off, organiser-run· IdrockBench Uzbek leaderboard data docs26.3%
9Gemma 4 E2B ITQ4_K_M, reasoning: off, organiser-run· IdrockBench Uzbek leaderboard data docs14.2%
10Nemotron 3.5 Lightning 30bQ4_K_M, reasoning: off, organiser-run· IdrockBench Uzbek leaderboard data docs11.6%
11Qwen3.5 2BQ8_0, reasoning: off, organiser-run· IdrockBench Uzbek leaderboard data docs7.9%
12Qwen3.5 0.8BQ8_0, reasoning: off, organiser-run· IdrockBench Uzbek leaderboard data docs1.4%

Data credits: IdrockBench, IDROCK AI Excellence Center, New Uzbekistan University. Public numeric results; no results licence stated (code MIT and dataset licences not borrowed).