Skip to content

Benchmark

Custom Voice · Public · Polish WER

Scores are percentages. Lower is better.

Models measured
1
Scores recorded
1
Sources
1
IQ weight
No weight published

Score distribution

1.5%
Mean
1.5%
Median
0%
Std dev
1.5%
10th percentile
1.5%
90th percentile
1.5%
Min

Leaderboard

#ModelScore
1Qwen3.5 Omni PlusTechnical report v2 · Table 12 · public multilingual set · Polish · WER· Reported by Qwen Team· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1.5%

Data credits: BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links.