Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Automatic Speech Recognition on Multilingual LibriSpeech (test)
Loading...
8.55
WER
Qwen3-ASR
7.744
13.1845
18.625
24.0655
Jan 29, 2026
WER
WER (Greedy)
Updated 4d ago
Evaluation Results
Method
Method
Links
WER
WER (Greedy)
Qwen3-ASR
Model Size=1.7B
2026.01
8.55
-
Whisper
Variant=large-v3
2026.01
8.62
-
Qwen3-ASR
Model Size=0.6B
2026.01
13.19
-
GLM-ASR
Variant=Nano-2512
2026.01
13.32
-
Fun-ASR
Variant=MLT-Nano
2026.01
28.7
-
Conformer-XL
GMACS=686, Training Se...
2023.05
-
5.32
FC-XL
GMACS=253, Training Se...
2023.05
-
4.9
FC-XXL
GMACS=441, Training Se...
2023.05
-
4.58
Feedback
Search any
task
Search any
task