Garble-Bench (Voice Arena)Models respond to dialogue with speech recognition errors. The evaluation measures error detection, transcription fidelity, conversation recovery and task completion.
Audio understanding and voice
Sep 2026
0.98
Not in cards
Voice Arena Full Duplex BenchVoice agents respond to pauses, changes of speaker, acknowledgments, and interruptions. The board averages success rates across these conversational behaviors.
Audio understanding and voice
99.9%
Not in cards
Voice Arena Diarization BenchModels identify who spoke when in recorded conversations. The board measures diarization error and publishes a separate track for speaker attributed transcription.
Audio understanding and voice
31%
Not in cards
Voice Arena Dictation BenchDictation tools turn speech into written drafts. Judges measure the share of drafts a user could send without edits.
Speech recognition and translation
75.8%
Not in cards
Monsoon ASRModels transcribe Indian English and Hindi recordings. The Hindi evaluation accounts for alternative spellings in its word error rate.
Speech recognition and translation
Jul 2026
Not in cards
Arenas without a leaderboard here
Name
Kind
Released
Best score
Status
Voice Arena Goal DatasetJudges compare two calls and choose which caller achieved the airline task more successfully. Human preferences provide the reference.