# Video Generation on VBench (Quality, Semantic, Total) Source: [Video Generation on VBench (Quality, Semantic, Total)]() Collection index: [llms.txt]() Data updated: 2026-06-29T07:18:17.960Z Evaluation of Video Generation models on the VBench dataset, focusing specifically on Quality, Semantic alignment, and the overall Total score. 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 | Video Generation | | Dataset (including any stated split) | VBench | | Metric | Quality Score | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 16 of 16 | | 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 - [Quality Score](): Higher is better; unit: PERCENT. - [Semantic Alignment Score](): Higher is better; unit: PERCENT. - [Total Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | Quality Score | Result date | Report paper | | --- | --- | --- | --- | | Causal-Forcing | 85.27 | Not recorded | 69bb83210636e5bf4afc1b4a | | AR-CoPO (semi) | 85.15 | Not recorded | 69bb83210636e5bf4afc1b4a | | LongLive | 85.1 | Not recorded | 69bb83210636e5bf4afc1b4a | | AR-CoPO (merged) | 85.07 | Not recorded | 69bb83210636e5bf4afc1b4a | | Self-Forcing | 84.87 | Not recorded | 69bb83210636e5bf4afc1b4a | | AR-CoPO (on-policy) | 84.81 | Not recorded | 69bb83210636e5bf4afc1b4a | | Casual Forcing | 84.02 | Not recorded | 6a41fbf3206542251d7cca47 | | Self-Forcing | 83.8 | Not recorded | 6a41fbf3206542251d7cca47 | | OTCA | 83.72 | Not recorded | 69e85bac7d0172140ec80feb | | DanceGRPO | 83.31 | Not recorded | 69e85bac7d0172140ec80feb | | Wan2.2 | 83.3 | Not recorded | 69e85bac7d0172140ec80feb | | Self-Forcing | 82.89 | Not recorded | 6a41fbf3206542251d7cca47 | | LongLive | 82.72 | Not recorded | 6a41fbf3206542251d7cca47 | | TempAct | 82.71 | Not recorded | 6a41fbf3206542251d7cca47 | | TempAct | 82.61 | Not recorded | 6a41fbf3206542251d7cca47 | | LongLive | 82.1 | Not recorded | 6a41fbf3206542251d7cca47 | ### Causal-Forcing Quality Score: 85.27. Result ID: 69bbc00547f34ce924a5e7ea. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### AR-CoPO (semi) Quality Score: 85.15. Result ID: 69bbc00547f34ce924a5e7ec. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"Self-Forcing","Alignment Strategy":"semi-on-policy"}. This record does not assert independent reproduction. ### LongLive Quality Score: 85.1. Result ID: 69bbc00547f34ce924a5e7eb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### AR-CoPO (merged) Quality Score: 85.07. Result ID: 69bbc00547f34ce924a5e7ee. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"Self-Forcing","Alignment Strategy":"semi-on-policy and on-policy merged"}. This record does not assert independent reproduction. ### Self-Forcing Quality Score: 84.87. Result ID: 69bbc00547f34ce924a5e7e9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### AR-CoPO (on-policy) Quality Score: 84.81. Result ID: 69bbc00547f34ce924a5e7ed. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"Self-Forcing","Alignment Strategy":"on-policy"}. This record does not assert independent reproduction. ### Casual Forcing Quality Score: 84.02. Result ID: 6a421c31fbc373592c872d35. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Prompting Strategy":"Single-prompt video generation"}. This record does not assert independent reproduction. ### Self-Forcing Quality Score: 83.8. Result ID: 6a421c31fbc373592c872d34. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Prompting Strategy":"Single-prompt video generation"}. This record does not assert independent reproduction. ### OTCA Quality Score: 83.72. Result ID: 69e9d5d647f34ce924d0e915. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### DanceGRPO Quality Score: 83.31. Result ID: 69e9d5d647f34ce924d0e914. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Wan2.2 Quality Score: 83.3. Result ID: 69e9d5d647f34ce924d0e913. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Self-Forcing Quality Score: 82.89. Result ID: 6a421c31fbc373592c872d37. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Prompting Strategy":"Step-prompt video generation"}. This record does not assert independent reproduction. ### LongLive Quality Score: 82.72. Result ID: 6a421c31fbc373592c872d36. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Prompting Strategy":"Single-prompt video generation"}. This record does not assert independent reproduction. ### TempAct Quality Score: 82.71. Result ID: 6a421c31fbc373592c872d38. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Prompting Strategy":"Step-prompt video generation","Backbone":"Self-Forcing"}. This record does not assert independent reproduction. ### TempAct Quality Score: 82.61. Result ID: 6a421c31fbc373592c872d3a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Prompting Strategy":"Step-prompt video generation","Backbone":"LongLive"}. This record does not assert independent reproduction. ### LongLive Quality Score: 82.1. Result ID: 6a421c31fbc373592c872d39. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Prompting Strategy":"Step-prompt video generation"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.