Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Joint ASR and Diarization on AliMeeting Far-field Mandarin (test)
Loading...
0.39
Error Rate
Qwen3.0-Omni-30B-A3B-Instruct
-0.8704
7.6373
16.145
24.6527
Jan 11, 2026
Error Rate
DER
cpCER
gCER
SCA
Updated 4d ago
Evaluation Results
Method
Method
Links
Error Rate
DER
cpCER
gCER
SCA
Qwen3.0-Omni-30B-A3B-Instruct
Model Type=End-to-End
2026.01
0.39
34.6
33.69
27.69
70.2
Pyannote 3.1 + Whisper-large-v3
Model Type=Cascade
2026.01
2.52
26.13
46.56
39.55
60.68
TagSpeech
Model Type=End-to-End
2026.01
3.04
22.13
33.84
25.42
81.63
Qwen2.5-Omni-7B
Model Type=End-to-End
2026.01
6.4
37.42
41.23
24.44
71.12
Gemini-2.0-flash
Model Type=End-to-End
2026.01
31.9
50.35
59.02
47.97
51.14
Feedback
Search any
task
Search any
task