Skip to content

Benchmark

Common Voice 15 · Cantonese WER/CER

Scores are percentages. Lower is better.

Models measured
2
Scores recorded
2
Sources
1
IQ weight
No weight published

Score distribution

2.7%
Mean
3.5%
Median
0.75%
Std dev
2%
10th percentile
3.5%
90th percentile
2%
Min
2%3.5%
The band covers the middle 80% of scores. The line marks the median.

Leaderboard

#ModelScore
1Qwen3.5 Omni PlusQwen3.5-Omni release · ASR WER/CER · lower is better · Cantonese· Reported by Qwen, Qwen Team· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.2%
2Qwen3.5 Omni FlashQwen3.5-Omni release · ASR WER/CER · lower is better · Cantonese· Reported by Qwen, Qwen Team· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.3.5%

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.