# Audio generation from Encodec tokens on universal evaluation dataset

Source: [Audio generation from Encodec tokens on universal evaluation dataset](<https://www.sota2.com/research/sota/audio-generation-from-encodec-tokens-on-universal-evaluation-dataset>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T05:55:02.609Z

Evaluation of audio generation models conditioned specifically on Encodec tokens, assessed using perceptual quality metrics like PESQ, ViSQOL, and MOS.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Audio generation from Encodec tokens |
| Dataset (including any stated split) | universal evaluation dataset |
| Metric | PESQ |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 27 of 27 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [PESQ](<https://www.sota2.com/research/sota/audio-generation-from-encodec-tokens-on-universal-evaluation-dataset.md?metric=PESQ>): Higher is better; unit: SCALAR.

- [ViSQOL](<https://www.sota2.com/research/sota/audio-generation-from-encodec-tokens-on-universal-evaluation-dataset.md?metric=ViSQOL>): Higher is better; unit: SCALAR.

- [FSD](<https://www.sota2.com/research/sota/audio-generation-from-encodec-tokens-on-universal-evaluation-dataset.md?metric=FSD>): Lower is better; unit: SCALAR.

- [SMOS](<https://www.sota2.com/research/sota/audio-generation-from-encodec-tokens-on-universal-evaluation-dataset.md?metric=SMOS>): Higher is better; unit: SCALAR.

- [MOS](<https://www.sota2.com/research/sota/audio-generation-from-encodec-tokens-on-universal-evaluation-dataset.md?metric=MOS>): Higher is better; unit: SCALAR.

## Recorded results

| Method | PESQ | Result date | Report paper |
| --- | --- | --- | --- |
| PeriodWave-Turbo | 3.579 | Not recorded | 69535c9453fdd5c36173da8f |
| Flow2GAN | 3.538 | Not recorded | 69535c9453fdd5c36173da8f |
| Flow2GAN | 3.457 | Not recorded | 69535c9453fdd5c36173da8f |
| Flow2GAN | 3.389 | Not recorded | 69535c9453fdd5c36173da8f |
| PeriodWave-Turbo | 3.229 | Not recorded | 69535c9453fdd5c36173da8f |
| RFWave | 3.117 | Not recorded | 69535c9453fdd5c36173da8f |
| Flow2GAN | 3.089 | Not recorded | 69535c9453fdd5c36173da8f |
| Flow2GAN | 2.983 | Not recorded | 69535c9453fdd5c36173da8f |
| Flow2GAN | 2.904 | Not recorded | 69535c9453fdd5c36173da8f |
| EnCodec | 2.776 | Not recorded | 69535c9453fdd5c36173da8f |
| RFWave | 2.663 | Not recorded | 69535c9453fdd5c36173da8f |
| Flow2GAN | 2.55 | Not recorded | 69535c9453fdd5c36173da8f |
| Flow2GAN | 2.442 | Not recorded | 69535c9453fdd5c36173da8f |
| Flow2GAN | 2.353 | Not recorded | 69535c9453fdd5c36173da8f |
| EnCodec | 2.259 | Not recorded | 69535c9453fdd5c36173da8f |
| PeriodWave-Turbo | 2.16 | Not recorded | 69535c9453fdd5c36173da8f |
| RFWave | 2.124 | Not recorded | 69535c9453fdd5c36173da8f |
| MBD | 2.094 | Not recorded | 69535c9453fdd5c36173da8f |
| MBD | 1.958 | Not recorded | 69535c9453fdd5c36173da8f |
| Flow2GAN | 1.925 | Not recorded | 69535c9453fdd5c36173da8f |
| Flow2GAN | 1.803 | Not recorded | 69535c9453fdd5c36173da8f |
| EnCodec | 1.745 | Not recorded | 69535c9453fdd5c36173da8f |
| Flow2GAN | 1.739 | Not recorded | 69535c9453fdd5c36173da8f |
| RFWave | 1.6 | Not recorded | 69535c9453fdd5c36173da8f |
| MBD | 1.457 | Not recorded | 69535c9453fdd5c36173da8f |
| EnCodec | 1.368 | Not recorded | 69535c9453fdd5c36173da8f |
| PeriodWave-Turbo | 1.26 | Not recorded | 69535c9453fdd5c36173da8f |

### PeriodWave-Turbo

PESQ: 3.579. Result ID: 699a178d5698665ec49bddc9.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74861e209131cea62328.md>)

[Implementation](<https://github.com/sh-lee-prml/PeriodWave>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"12.0","Parameters":"31.3M","Sampling Steps":4}. This record does not assert independent reproduction.

### Flow2GAN

PESQ: 3.538. Result ID: 699a178d5698665ec49bddcc.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"12.0","Parameters":"78.9M","Sampling Steps":4}. This record does not assert independent reproduction.

### Flow2GAN

PESQ: 3.457. Result ID: 699a178d5698665ec49bddcb.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"12.0","Parameters":"78.9M","Sampling Steps":2}. This record does not assert independent reproduction.

### Flow2GAN

PESQ: 3.389. Result ID: 699a178d5698665ec49bddca.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"12.0","Parameters":"78.9M","Sampling Steps":1}. This record does not assert independent reproduction.

### PeriodWave-Turbo

PESQ: 3.229. Result ID: 699a178d5698665ec49bddc3.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74861e209131cea62328.md>)

[Implementation](<https://github.com/sh-lee-prml/PeriodWave>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"6.0","Parameters":"31.3M","Sampling Steps":4}. This record does not assert independent reproduction.

### RFWave

PESQ: 3.117. Result ID: 699a178d5698665ec49bddc8.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75111e209131cea657ba.md>)

[Implementation](<https://github.com/bfs18/rfwave>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"12.0","Parameters":"18.1M","Sampling Steps":10}. This record does not assert independent reproduction.

### Flow2GAN

PESQ: 3.089. Result ID: 699a178d5698665ec49bddc6.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"6.0","Parameters":"78.9M","Sampling Steps":4}. This record does not assert independent reproduction.

### Flow2GAN

PESQ: 2.983. Result ID: 699a178d5698665ec49bddc5.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"6.0","Parameters":"78.9M","Sampling Steps":2}. This record does not assert independent reproduction.

### Flow2GAN

PESQ: 2.904. Result ID: 699a178d5698665ec49bddc4.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"6.0","Parameters":"78.9M","Sampling Steps":1}. This record does not assert independent reproduction.

### EnCodec

PESQ: 2.776. Result ID: 699a178d5698665ec49bddc7.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6db91e209131cea3b969.md>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"12.0","Parameters":"56.0M"}. This record does not assert independent reproduction.

### RFWave

PESQ: 2.663. Result ID: 699a178d5698665ec49bddc2.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75111e209131cea657ba.md>)

[Implementation](<https://github.com/bfs18/rfwave>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"6.0","Parameters":"18.1M","Sampling Steps":10}. This record does not assert independent reproduction.

### Flow2GAN

PESQ: 2.55. Result ID: 699a178d5698665ec49bddbf.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"3.0","Parameters":"78.9M","Sampling Steps":4}. This record does not assert independent reproduction.

### Flow2GAN

PESQ: 2.442. Result ID: 699a178d5698665ec49bddbe.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"3.0","Parameters":"78.9M","Sampling Steps":2}. This record does not assert independent reproduction.

### Flow2GAN

PESQ: 2.353. Result ID: 699a178d5698665ec49bddbd.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"3.0","Parameters":"78.9M","Sampling Steps":1}. This record does not assert independent reproduction.

### EnCodec

PESQ: 2.259. Result ID: 699a178d5698665ec49bddc0.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6db91e209131cea3b969.md>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"6.0","Parameters":"56.0M"}. This record does not assert independent reproduction.

### PeriodWave-Turbo

PESQ: 2.16. Result ID: 699a178d5698665ec49bddbc.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74861e209131cea62328.md>)

[Implementation](<https://github.com/sh-lee-prml/PeriodWave>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"3.0","Parameters":"31.3M","Sampling Steps":4}. This record does not assert independent reproduction.

### RFWave

PESQ: 2.124. Result ID: 699a178d5698665ec49bddbb.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75111e209131cea657ba.md>)

[Implementation](<https://github.com/bfs18/rfwave>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"3.0","Parameters":"18.1M","Sampling Steps":10}. This record does not assert independent reproduction.

### MBD

PESQ: 2.094. Result ID: 699a178d5698665ec49bddc1.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa711d1e209131cea4f2ef.md>)

[Implementation](<https://github.com/facebookresearch/audiocraft>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"6.0","Parameters":"411.0M"}. This record does not assert independent reproduction.

### MBD

PESQ: 1.958. Result ID: 699a178d5698665ec49bddba.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa711d1e209131cea4f2ef.md>)

[Implementation](<https://github.com/facebookresearch/audiocraft>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"3.0","Parameters":"411.0M"}. This record does not assert independent reproduction.

### Flow2GAN

PESQ: 1.925. Result ID: 699a178d5698665ec49bddb8.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"1.5","Parameters":"78.9M","Sampling Steps":4}. This record does not assert independent reproduction.

### Flow2GAN

PESQ: 1.803. Result ID: 699a178d5698665ec49bddb7.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"1.5","Parameters":"78.9M","Sampling Steps":2}. This record does not assert independent reproduction.

### EnCodec

PESQ: 1.745. Result ID: 699a178d5698665ec49bddb9.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6db91e209131cea3b969.md>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"3.0","Parameters":"56.0M"}. This record does not assert independent reproduction.

### Flow2GAN

PESQ: 1.739. Result ID: 699a178d5698665ec49bddb6.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"1.5","Parameters":"78.9M","Sampling Steps":1}. This record does not assert independent reproduction.

### RFWave

PESQ: 1.6. Result ID: 699a178d5698665ec49bddb4.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75111e209131cea657ba.md>)

[Implementation](<https://github.com/bfs18/rfwave>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"1.5","Parameters":"18.1M","Sampling Steps":10}. This record does not assert independent reproduction.

### MBD

PESQ: 1.457. Result ID: 699a178d5698665ec49bddb3.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa711d1e209131cea4f2ef.md>)

[Implementation](<https://github.com/facebookresearch/audiocraft>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"1.5","Parameters":"411.0M"}. This record does not assert independent reproduction.

### EnCodec

PESQ: 1.368. Result ID: 699a178d5698665ec49bddb2.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6db91e209131cea3b969.md>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"1.5","Parameters":"56.0M"}. This record does not assert independent reproduction.

### PeriodWave-Turbo

PESQ: 1.26. Result ID: 699a178d5698665ec49bddb5.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9453fdd5c36173da8f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74861e209131cea62328.md>)

[Implementation](<https://github.com/sh-lee-prml/PeriodWave>)

Recorded evaluation configuration: {"Bandwidth (kbps)":"1.5","Parameters":"31.3M","Sampling Steps":4}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.
