Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

VoiceBench

Benchmarks

Task NameDataset NameSOTA ResultTrend
General Audio UnderstandingVoiceBench
AlpacaEval Score4.78
19
Speech-to-TextVoiceBench
AlpacaEval Score4.78
15
Speech-to-text reasoning and semantic understandingVoiceBench (test)
Alpaca Eval4.8
13
ReasoningVoiceBench
MMSU Accuracy (Audio)72.9
13
Voice EvaluationVoiceBench
Overall Score (VoiceBench)89.6
10
Audio Instruction FollowingVoiceBench
AlpacaEval Score4.78
10
Spoken Question AnsweringVoiceBench
Accuracy76.79
9
General capability evaluationVoicebench
HS Score76.91
8
Empathetic Speech GenerationVoiceBench CommonEval
Empathy Score4.22
7
Speech-to-Text Spoken Question AnsweringVoiceBench S2T (test)
AlpacaEval4.8
7
Voice ChattingVoiceBench
AlpacaEval4.94
7
Conversational IntelligenceVoiceBench
AlpacaEval4.57
6
General ConversationVoiceBench
AlpacaEval Score4.19
5
Spoken Question AnsweringVoiceBench S2T
AlpacaEval4.94
4
Speech Question AnsweringVoiceBench AlpacaEval
AlpacaEval Score4.81
3
Voice InteractionVoiceBench
VoiceBench Average Score89.4
3
DialogueVoiceBench
Score93.1
3
Showing 17 of 17 rows