Audio Reconstruction on LibriSpeech clean (test)
1STOIGround Truth
Evaluation Results
| Method | Links | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Ground TruthBitrate (kbps)=-, Semantic Info.=-2026.02 | 1 | 4.64 | 5 | 1 | — | — | — | — | — | — | — | — | |
| DAC-24kSampling Rate=24 kHz, fr (Hz)=75, nq=32, |V|=1024, R (bps)=240002026.06 | 0.99 | — | — | 0.94 | — | — | — | 1.02 | — | — | 5.01 | 4.06 | |
| STACodecBitrate (kbps)=4.0, Semantic Info.=HuBERT-base2026.02 | 0.97 | 3.61 | 4.5 | 0.93 | — | — | — | — | — | — | — | — | |
| STACodecBitrate (kbps)=4.0, Semantic Info.=WavLM-large2026.02 | 0.97 | 3.62 | 4.51 | 0.93 | — | — | — | — | — | — | — | — | |
| DAC-16kSampling Rate=16 kHz, fr (Hz)=50, nq=12, |V|=1024, R (bps)=60002026.06 | 0.97 | — | — | 0.9 | — | — | — | 2.04 | — | — | 5.25 | 4 | |
| PASTBitrate (kbps)=4.0, Semantic Info.=ASR and PR2026.02 | 0.96 | 3.16 | 4.32 | 0.86 | — | — | — | — | — | — | — | — | |
| STACodec-SPDBitrate (kbps)=4.0, Semantic Info.=WavLM-large2026.02 | 0.96 | 3.51 | 4.43 | 0.92 | — | — | — | — | — | — | — | — | |
| OmniCodec-32LCodebook Size=2048, Bitrate (bps)=4400, TPS=12.5×322026.03 | 0.96 | — | — | — | 3.02 | 3.62 | 0.75 | 2.54 | 3.63 | 3.73 | — | — | |
| MimiSampling Rate=24 kHz, fr (Hz)=12.5, nq=8, |V|=2048, R (bps)=11002026.06 | 0.96 | — | — | 0.88 | — | — | — | 2.49 | — | — | 5.52 | 3.92 | |
| Mimi codec-16LCodebook Size=2048, Bitrate (bps)=2200, TPS=12.5×162026.03 | 0.94 | — | — | — | 2.88 | 3.42 | 1.12 | 3.93 | 3.42 | 3.51 | — | — | |
| OmniCodec-16LCodebook Size=2048, Bitrate (bps)=2200, TPS=12.5×162026.03 | 0.94 | — | — | — | 2.76 | 3.41 | 0.81 | 3.05 | 3.45 | 3.5 | — | — | |
| X-Codec-8LCodebook Size=1024, Bitrate (bps)=4000, TPS=50×82026.03 | 0.93 | — | — | — | 2.96 | 3.49 | 0.79 | 3.35 | 3.56 | 3.67 | — | — | |
| OmniCodec-F-32LCodebook Size=2048, Bitrate (bps)=2200, TPS=6.25×322026.03 | 0.93 | — | — | — | 2.52 | 3.19 | 0.84 | 3.15 | 3.41 | 3.47 | — | — | |
| BigCodecSampling Rate=16 kHz, fr (Hz)=80, nq=1, |V|=8192, R (bps)=10402026.06 | 0.93 | — | — | 0.81 | — | — | — | 3.01 | — | — | 7.77 | 4.11 | |
| SpeechTokenizerBitrate (kbps)=4.0, Semantic Info.=HuBERT-base2026.02 | 0.92 | 2.6 | 4.26 | 0.85 | — | — | — | — | — | — | — | — | |
| UniCodecCodebook Size=16384, Bitrate (bps)=1050, TPS=752026.03 | 0.92 | — | — | — | 2.65 | 3.14 | 0.79 | 3.46 | 3.2 | 3.31 | — | — | |
| AUVCodebook Size=20480, Bitrate (bps)=716, TPS=502026.03 | 0.91 | — | — | — | 2.4 | 3.05 | 0.89 | 3.6 | 3.15 | 3.23 | — | — | |
| OmniCodec-8LCodebook Size=2048, Bitrate (bps)=1100, TPS=12.5×82026.03 | 0.91 | — | — | — | 2.14 | 2.78 | 0.9 | 3.78 | 3.26 | 3.29 | — | — | |
| OmniCodec-F-16LCodebook Size=2048, Bitrate (bps)=1100, TPS=6.25×162026.03 | 0.9 | — | — | — | 2.14 | 2.79 | 0.89 | 3.83 | 3.21 | 3.27 | — | — | |
| WavTokenizer-75Sampling Rate=24 kHz, fr (Hz)=75, nq=1, |V|=4096, R (bps)=9002026.06 | 0.9 | — | — | 0.66 | — | — | — | 3.92 | — | — | 11.37 | 3.79 | |
| X-CodecBitrate (kbps)=4.0, Semantic Info.=HuBERT-base2026.02 | 0.88 | 2.79 | 4.27 | 0.88 | — | — | — | — | — | — | — | — | |
| SNACSampling Rate=24 kHz, fr (Hz)=12,23,47, nq=3, |V|=4096, R (bps)=9842026.06 | 0.88 | — | — | 0.59 | — | — | — | 4.07 | — | — | 13.14 | 3.02 | |
| WavTokenizerCodebook Size=4096, Bitrate (bps)=480, TPS=402026.03 | 0.87 | — | — | — | 1.88 | 2.44 | 0.97 | 4.79 | 3.02 | 3.11 | — | — | |
| WavTokenizer-40Sampling Rate=24 kHz, fr (Hz)=40, nq=1, |V|=4096, R (bps)=4802026.06 | 0.85 | — | — | 0.52 | — | — | — | 4.99 | — | — | 24.36 | 3.57 | |
| Qwen3-TTS-Tok.Sampling Rate=24 kHz, fr (Hz)=12.5, nq=16, |V|=2048, R (bps)=22002026.06 | 0.65 | — | — | 0.91 | — | — | — | 7.11 | — | — | 5.43 | 4.13 | |
| HASRDBitrate (kbps)=4.5, Semantic Info.=HuBERT-base2026.02 | — | — | 4.3 | — | — | — | — | — | — | — | — | — | |
| HASRDBitrate (kbps)=3.1, Semantic Info.=BestRQ+2026.02 | — | — | 4.5 | — | — | — | — | — | — | — | — | — |