# Text-to-Music Generation on MusicBench (test) Source: [Text-to-Music Generation on MusicBench (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:59:47.314Z Evaluation of Text-to-Music Generation models using the MusicBench test set, assessed via human preference metrics across general audience and music expert panels. 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 | Text-to-Music Generation | | Dataset (including any stated split) | MusicBench (test) | | Metric | REL (General) | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 6 of 6 | | 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 - [REL (General)](): Higher is better; unit: SCALAR. - [AQ (General)](): Higher is better; unit: SCALAR. - [OMQ (General)](): Higher is better; unit: SCALAR. - [RC (General)](): Higher is better; unit: SCALAR. - [HC (General)](): Higher is better; unit: SCALAR. - [REL (Expert)](): Higher is better; unit: SCALAR. - [MCM (Expert)](): Higher is better; unit: SCALAR. - [MTM (Expert)](): Higher is better; unit: SCALAR. - [AQ (Expert)](): Higher is better; unit: SCALAR. - [OMQ (Expert)](): Higher is better; unit: SCALAR. - [RC (Expert)](): Higher is better; unit: SCALAR. - [HC (Expert)](): Higher is better; unit: SCALAR. ## Recorded results | Method | REL (General) | Result date | Report paper | | --- | --- | --- | --- | | Mustango | 5.18 | Not recorded | 68fa805b1e209131cea9cc5d | | Tango | 4.96 | Not recorded | 68fa805b1e209131cea9cc5d | | Mustango | 4.85 | Not recorded | 68fa805b1e209131cea9cc5d | | Mustango | 4.79 | Not recorded | 68fa805b1e209131cea9cc5d | | MusicGen-M | 4.55 | Not recorded | 68fa805b1e209131cea9cc5d | | AudioLDM2 | 3.99 | Not recorded | 68fa805b1e209131cea9cc5d | ### Mustango REL (General): 5.18. Result ID: 698b95bc7a967eddb76fcc10. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pre-trained":false,"Study Round":"Second round"}. This record does not assert independent reproduction. ### Tango REL (General): 4.96. Result ID: 698b95bc7a967eddb76fcc0b. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pre-trained":true,"Study Round":"First run"}. This record does not assert independent reproduction. ### Mustango REL (General): 4.85. Result ID: 698b95bc7a967eddb76fcc0c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pre-trained":true,"Study Round":"First run"}. This record does not assert independent reproduction. ### Mustango REL (General): 4.79. Result ID: 698b95bc7a967eddb76fcc0d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pre-trained":false,"Study Round":"First run"}. This record does not assert independent reproduction. ### MusicGen-M REL (General): 4.55. Result ID: 698b95bc7a967eddb76fcc0e. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Study Round":"Second round"}. This record does not assert independent reproduction. ### AudioLDM2 REL (General): 3.99. Result ID: 698b95bc7a967eddb76fcc0f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Study Round":"Second round"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.