Skip to content

Benchmark

UzLiB fill-in accuracy

Multilingual benchmarks compare model performance across languages. Higher scores are better.

Models measured
74
Scores recorded
74
Sources
1
IQ weight
1.36

Score distribution

0.449
Mean
0.423
Median
0.156
Std dev
0.269
10th percentile
0.692
90th percentile
0.173
Min
0.1730.808
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.36 for chat models, measured over 73 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51GLM 4.6temperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.346
52GPT-5 minitemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.346
53GPT-5 nanotemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.346
54MiMo V2 Flashtemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.346
55Qwen 3.8 27Btemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.346
56Claude Sonnet 4.5temperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.327
57Kimi K2 Instructtemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.327
58Llama 4 Mavericktemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.327
59Qwen3 4Btemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.327
60Gemini 2.5 Flash-Litetemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.308
61Gemma 3 27B ITtemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.308
62GLM 5.2temperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.308
63Mistral Nemotemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.308
64Llama 3.3 70B Instructtemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.289
65Llama 4 Scouttemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.289
66Llama 3.2 1B Instructtemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.269
67Qwen3 14Btemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.269
68Qwen3.5 9Btemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.269
69Gemma 3n E4B ITtemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.25
70Gemma 3 12B ITtemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.231
71GPT OSS 20btemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.231
72Mistral (7B) Instruct v0.3temperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.192
73Llama 3.1 8B Instructtemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.173
74Llama 3.2 3Btemperature: 1.0, top_p: 0.95· UzLiB Uzbek linguistic benchmark data docs0.173

Showing 51–74 of 74 results.

Data credits: UzLiB, Tahrirchi (Abror Shopulatov). Public LEADERBOARD.md numbers; repository MIT covers software and documentation, no separate results licence stated and none asserted.