Subscribe
Benchmarks
verified 2026-11. Download on Hugging Face
34 of 93
SaveFull-duplexTierSettingFocus
Full-Duplex-Bench v1NTU / UC Berkeley / UW / MIT
Full-duplexSpeech-nativeLab—142025-03
DuplexActDong et al.
Full-duplexSpeech-nativeLab—82026-10
Full-Duplex-Bench v1.5NTU / UC Berkeley / UW / MIT
Full-duplexSpeech-nativeLab—62025-08
Full-Duplex-Bench v3NTU (NVIDIA advisory)
Full-duplexSpeech-nativeLiveVertical agent62026-05
Full-duplexSpeech-nativeVerticalVertical agent Speech-LM52026-10
VideoFDBNVIDIA / David AI
Full-duplexSpeech-nativeLive—42026-10
Audio MultiChallengeScale AI (Scale Labs)
Full-duplexSpeech-nativeLabSpeech-LM42026-10
TurnBenchSesame AI Labs / Mundo AI / CMU / oto
Full-duplexSpeech-nativeLab—32026-11
Talking TurnsApple / CMU
Full-duplexSpeech-nativeLabSpeech-LM32025-04
Full-Duplex-Bench v2NTU / UC Berkeley / UW / MIT
Full-duplexSpeech-nativeLiveSpeech-LM22026-02
MultiPartyTurnBenchMathur and Manocha
Full-duplexSpeech-nativeLab—22026-10
M3-DuplexBenchNTT (Fukuda et al.)
Full-duplexSpeech-nativeLabSpeech-LM22026-10
HumDial (ICASSP '26)ICASSP 2026 Spoken Dialogue Challenge
Full-duplexSpeech-nativeLabSpeech-LM Emotion12026-01
J-Moshi (subjective)Ohashi et al.
Full-duplexSpeech-nativeLabSpeech-LM12025-06
MultiTalkBenchWang et al. (MultiTalk)
Full-duplexSpeech-nativeLab—02026-09
Duplex-MPEMa et al.
Full-duplexSpeech-nativeLab—02026-11
DuplexLedgerLi et al.
Full-duplexSpeech-nativeLabSpeech-LM02026-10
InteractionBenchAvera Labs (He et al.)
Full-duplexSpeech-nativeLiveSpeech-LM02026-11
ECHOZhao et al.
Full-duplexSpeech-nativeLab—02026-09
MP-BenchShih et al.
Full-duplexSpeech-nativeLabSpeech-LM02026-10
DuplexSpeechBench-IFEvalMathur and Manocha
Full-duplexSpeech-nativeLabInstruction following02026-09
APEX-VoiceMathur and Manocha
Full-duplexSpeech-nativeVerticalVertical agent02026-09
SID-BenchXia et al.
Full-duplexSpeech-nativeLab—02026-03
FD-BenchPeng et al.
Full-duplexSpeech-nativeLab—02025-07
MTR-DuplexBenchZhang et al.
Full-duplexSpeech-nativeLabSpeech-LM02026-04
FLEXIGe et al.
Full-duplexSpeech-nativeLab—02025-09
MMedFDAcademic (MMedFD authors)
Full-duplexAdjacentVerticalASR Vertical agent02025-09
Omni-DuplexEvalOpenBMB (Tsinghua University / Shanghai Qi Zhi Institute / ModelBest)
Full-duplexAdjacentLabAudio understanding02026-07
OmniInteractCUHK MMLab / SJTU / NTU / McMaster / CityUHK / JUFE
Full-duplexSpeech-nativeLiveAudio understanding02026-05
eot-benchLiveKit
Full-duplexSpeech-nativeLab—02026-11
Full-duplexSpeech-nativeLive—02026-11
CoDeTTBairong (BRVoice team)
Full-duplexSpeech-nativeLab—02026-04
DuplexWorldCentific / University of Maryland
Full-duplexSpeech-nativeVerticalVertical agent02026-08
Game-Time BenchmarkMIT / NTU / Academia Sinica
Full-duplexSpeech-nativeLabInstruction following02026-06