Skip to content

Benchmark

LaoBench Lao-500 judge difference Δ(Gemini-2.5-Pro − Qwen3-Max)

Multilingual benchmarks compare model performance across languages.

Models and systems measured
11
Scores recorded
12
Sources
1
IQ weight
No weight published

Score distribution

-0.254
Mean
0.78
Median
3.7
Std dev
-6.2
10th percentile
3.3
90th percentile
-7.6
Min
-7.65.4
The band covers the middle 80% of scores. The line marks the median.

Results

The source marks these results as descriptive, not capability scores, so we list them by name instead of ranking them.

Model or systemScore
Claude Opus 4.1table: 10, statistic: delta, mode: single· LaoBench (arXiv:2511.11334v3) data3.3
Claude Sonnet 4.5table: 10, statistic: delta, mode: single· LaoBench (arXiv:2511.11334v3) data0.42
DeepSeek V3.2 Exptable: 10, statistic: delta, mode: Non-Thinking· LaoBench (arXiv:2511.11334v3) data1.6
DeepSeek V3.2 Exptable: 10, statistic: delta, mode: Thinking· LaoBench (arXiv:2511.11334v3) data1.4
Gemini 2.5 Protable: 10, statistic: delta, mode: single· LaoBench (arXiv:2511.11334v3) data5.4
GPT 5table: 10, statistic: delta, mode: single-self-baseline· LaoBench (arXiv:2511.11334v3) data0
Qwen 3 235B A22Btable: 10, statistic: delta, mode: single· LaoBench (arXiv:2511.11334v3) data0.78
Qwen3 235B A22B Instruct 2507table: 10, statistic: delta, mode: single· LaoBench (arXiv:2511.11334v3) data-6.2
Qwen3 Maxtable: 10, statistic: delta, mode: single· LaoBench (arXiv:2511.11334v3) data-7.6
Qwen3 Next 80B A3B Instructtable: 10, statistic: delta, mode: single· LaoBench (arXiv:2511.11334v3) data-1.2
Qwen3-PlusNon-Thinkingtable: 10, statistic: delta, mode: Non-Thinking· LaoBench (arXiv:2511.11334v3) data-3.4
Qwen3-PlusThinkingtable: 10, statistic: delta, mode: Thinking· LaoBench (arXiv:2511.11334v3) data2.5

Data credits: LaoBench, BAAI et al., arXiv:2511.11334v3. Cited numeric results only. arXiv non-exclusive distribution notice establishes no results reuse grant; dataset Apache licence is not a results grant.