Skip to content

Benchmark

ArmBench-LLM Armenian exam_history (armenian_exam_score, 0-shot)

Higher scores are better.

Models measured
27
Scores recorded
28
Sources
1
IQ weight
1.34

Score distribution

6.7
Mean
6.3
Median
3.7
Std dev
2.8
10th percentile
12.8
90th percentile
1
Min
114.5
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.34 for chat models, measured over 27 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Gemini 3 Pro Preview· ArmBench-LLM Armenian data docs14.5
2Gemini 3 Flash· ArmBench-LLM Armenian data docs13.3
3GPT-5.2 Prosetting not named by the source· ArmBench-LLM Armenian data docs12.8
4GPT-5.2 Prosetting not named by the source· ArmBench-LLM Armenian data docs12.8
5Gemini 2.5 Pro· ArmBench-LLM Armenian data docs11.3
6Gemini 3.1 Pro Preview· ArmBench-LLM Armenian data docs10.5
7Gemini 2.5 Flash· ArmBench-LLM Armenian data docs9.5
8Grok 4.1 Fast· ArmBench-LLM Armenian data docs8.5
9GLM 5· ArmBench-LLM Armenian data docs8.3
10Claude Sonnet 4· ArmBench-LLM Armenian data docs7.5
11Mistral Large 3· ArmBench-LLM Armenian data docs7.5
12GPT-5.4 mini· ArmBench-LLM Armenian data docs7
13Claude Sonnet 3.7· ArmBench-LLM Armenian data docs6.8
14Grok 4 Fast· ArmBench-LLM Armenian data docs6.3
15Qwen3.5 27B· ArmBench-LLM Armenian data docs6.3
16Gemma 4 31B IT· ArmBench-LLM Armenian data docs5.5
17Qwen3.5 35B A3B· ArmBench-LLM Armenian data docs4.5
18Grok 4.20· ArmBench-LLM Armenian data docs4.3
19Gemma 3 4B IT· ArmBench-LLM Armenian data docs4
20Mistral Small 3.2 24B Instruct· ArmBench-LLM Armenian data docs4
21Gemma 3 12B IT· ArmBench-LLM Armenian data docs3.8
22GPT 5.4 Nano· ArmBench-LLM Armenian data docs3.8
23Ministral 14B· ArmBench-LLM Armenian data docs3.8
24Gemma 3 27B IT· ArmBench-LLM Armenian data docs3.3
25Qwen3 4B Instruct 2507· ArmBench-LLM Armenian data docs2.8
26Qwen3 4B Base· ArmBench-LLM Armenian data docs2
27GLM 4.7 Flash· ArmBench-LLM Armenian data docs1.8
28Qwen3.5 9B· ArmBench-LLM Armenian data docs1

Data credits: ArmBench-LLM by Metric-AI. Public numeric primary task results; no separate results licence asserted.