List your company

Speech Reconstruction on LibriTTS (test-other)

4.17UTMOSCleanCodec@12.5
1.3917442.1130222.83433.555578Feb 27, 2025May 14, 2025Jul 30, 2025Oct 15, 2025Dec 31, 2025Mar 18, 2026Jun 3, 2026

Evaluation results

MethodLinks
CleanCodec@12.5Token Rate (t/s)=12.52026.064.17————5.7—3.20.8421.730.44
CleanCodec@31.25Token Rate (t/s)=31.252026.064.1————3.6—1.90.8722.120.33
KanadeToken Rate (t/s)=12.52026.064.1————8.2—4.50.6423.770.59
Qwen3Token Rate (t/s)=2002026.064.09————2.6—1.40.927.850.29
BiCodecToken Rate (t/s)=502026.064.07————6.2—3.40.7726.140.62
CleanCodec@62.5Token Rate (t/s)=62.52026.064.06————2.7—1.60.8922.830.29
FocalCodecToken Rate (t/s)=12.52026.064.05————16.8—9.70.4225.250.79
GTToken Rate (t/s)=–2026.064.01————0—0124.440
XCodec2Token Rate (t/s)=502026.064.01————5.4—3.10.7523.470.4
SylberBitrate (Hz)=4.222026.013.911.130.72——13.290.71————
Qwen3Token Rate (t/s)=1002026.063.73————4.7—2.50.7220.20.38
TAAEUnified=true, DR=320, TPF=2, TPS=502025.023.71361.75390.8380.8994———————
MimiToken Rate (t/s)=1002026.063.63————7.5—4.40.717.210.56
WavTokenizerToken Rate (t/s)=402026.063.63————19.1—11.50.6120.940.47
UniCodecUnified=true, DR=320, TPF=1, TPS=752025.023.582.27490.90950.9109———————
BigCodecUnified=false, DR=200, TPF=1, TPS=802025.023.54532.38170.90940.9237———————
Sylber 2.0Bitrate (Hz)=5.812026.013.541.890.87——8.580.91————
Q2D2Bandwidth=1kbps, Nq=1, token/s=752025.123.53832.22240.8908—0.9199——————
Q2D2Bandwidth=3.3kbps, Nq=1, token/s=1662025.123.50723.0960.9339—0.9531——————
Ground Truth2025.023.4831——————————
GT2025.123.4831——————————
FACodecBitrate (Hz)=802026.013.482.610.93——7.140.96————
SACodecParams [M]=75.17, Token Rate=75, Codebook Size=1000 / 1024, Q=2, Bitrate [kbps]=1.52025.123.47862.40160.9040.9273———————
Q2D2Bandwidth=6.9kbps, Nq=1, token/s=3332025.123.44813.45950.9464—0.9712——————
WavTokenizerParams [M]=80.55, Token Rate=75, Codebook Size=4096, Q=1, Bitrate [kbps]=0.92025.123.43152.27050.91730.8907———————
Wavtokenizer (speech)Unified=false, DR=320, TPF=1, TPS=752025.023.43122.26140.89070.9172———————
WavTokenizerBandwidth=0.9kbps, Nq=1, token/s=752025.123.43122.26140.8907—0.9172——————
DACBitrate (Hz)=86.12026.013.44.430.99——5.991————
DACUnified=false, DR=48.9, TPF=9, TPS=9002025.023.35663.75950.95760.9696———————
DACBandwidth=9.0kbps, Nq=9, token/s=9002025.123.35663.75950.9576—0.9696——————
MimiBitrate (Hz)=12.52026.013.333.250.95——7.060.97————
HiFi-CodecBandwidth=3.0kbps, Nq=4, token/s=3002025.123.30342.60830.9166—0.9318——————
SpeechTokenizerUnified=false, DR=40, TPF=8, TPS=6002025.023.28512.32690.88110.9205———————
SpeechTokenizerParams [M]=103.68, Token Rate=50, Codebook Size=1024, Q=8, Bitrate [kbps]=42025.123.28512.32690.88110.9205———————
SpeechTokenizerBandwidth=6.0kbps, Nq=8, token/s=6002025.123.28512.32690.8811—0.9205——————
SpeechTokenizerBitrate (Hz)=502026.013.282.410.9——7.810.95————
VocosBandwidth=6.0kbps, Nq=8, token/s=6002025.123.19562.5590.9209—0.9202——————
DACParams [M]=74.71, Token Rate=75, Codebook Size=1024, Q=8, Bitrate [kbps]=62025.123.13383.34290.94020.9598———————
Wavtokenizer (unified)Unified=true, DR=320, TPF=1, TPS=752025.023.0821.66490.83120.8874———————
WavTokenizerBitrate (Hz)=402026.013.081.690.84——31.040.87————
HiFi-CodecBandwidth=4.0kbps, Nq=4, token/s=4002025.123.0752.55360.9126—0.9387——————
MimiUnified=true, DR=240, TPF=8, TPS=1002025.023.06082.09520.88160.8875———————
VocosBandwidth=3.0kbps, Nq=4, token/s=3002025.123.05582.19330.8967—0.9051——————
WavTokenizerBandwidth=0.5kbps, Nq=1, token/s=402025.123.05451.66220.8336—0.8953——————
SpeechTokenizerBandwidth=3.0kbps, Nq=4, token/s=3002025.123.01831.73730.8371—0.8907——————
DACBandwidth=4.0kbps, Nq=4, token/s=4002025.122.94482.59480.9083—0.9404——————
FACodecParams [M]=374.49, Token Rate=80, Codebook Size=1024, Q=6, Bitrate [kbps]=4.82025.122.93022.03210.88320.908———————
EncodecUnified=true, DR=40, TPF=8, TPS=6002025.022.65682.68180.92410.9338———————
EncodecParams [M]=14.85, Token Rate=75, Codebook Size=1024, Q=8, Bitrate [kbps]=62025.122.65682.68180.92410.9338———————
EncodecBandwidth=6.0kbps, Nq=8, token/s=6002025.122.65682.68180.9241—0.9338——————
Qwen3Token Rate (t/s)=502026.062.51————15.8—9.40.38.910.61
EncodecBandwidth=3.0kbps, Nq=4, token/s=3002025.122.08832.0520.8835—0.8926——————
SpeechTokenizerParams [M]=103.68, Token Rate=50, Codebook Size=1024, Q=3, Bitrate [kbps]=1.52025.122.01041.22410.7780.7445———————
DACParams [M]=74.71, Token Rate=75, Codebook Size=1024, Q=2, Bitrate [kbps]=1.52025.121.74431.50390.82180.8636———————
EncodecParams [M]=14.85, Token Rate=75, Codebook Size=1024, Q=2, Bitrate [kbps]=1.52025.121.51321.57530.82910.8228———————
DACUnified=false, DR=440, TPF=1, TPS=1002025.021.49861.24540.75050.7775———————
DACBandwidth=1.0kbps, Nq=1, token/s=1002025.121.49861.24540.7505—0.7775——————