VBench
Video generation models make clips from prompts, and automatic checks score 16 dimensions of quality and prompt adherence.
Results give Quality, Semantic and Total scores. VBench++ (TPAMI 2025) extends it to image-to-video and trustworthiness. VBench-2.0 (2025-03) tests intrinsic faithfulness with new dimensions. VBench and V*Bench are different tests.
- Made by
- NTU S-Lab and Shanghai AI Laboratory (Vchitect)
- Released
- November 2023
- Kind
- Video generation
- Level
- Everyday
- Scored
- Checked automatically
- Answers
- Open-ended
- Tasks
- All public
- Last in a launch post
- February 2026, dropped after it saturated
- Last model added
- February 2026
- Scores refreshed
- November 12th, 2026
In the authors’ words
Video generation has witnessed significant advancements, yet evaluating these models remains a challenge. A comprehensive evaluation benchmark for video generation is indispensable for two reasons: 1) Existing metrics do not fully align with human perceptions; 2) An ideal evaluation system should provide insights to inform future developments of video generation. To this end, we present VBench, a comprehensive benchmark suite that dissects "video generation quality" into specific, hierarchical, and disentangled dimensions, each with tailored prompts and evaluation methods. VBench has three appealing properties: 1) Comprehensive Dimensions: VBench comprises 16 dimensions in video generation (e.g., subject identity inconsistency, motion smoothness, temporal flickering, and spatial relationship, etc). The evaluation metrics with fine-grained levels reveal individual models' strengths and weaknesses. 2) Human Alignment: We also provide a dataset of human preference annotations to validate our benchmarks' alignment with human perception, for each evaluation dimension respectively. 3) Valuable Insights: We look into current models' ability across various evaluation dimensions, and various content types. We also investigate the gaps between video and image generation models. We will open-source VBench, including all prompts, evaluation methods, generated videos, and human preference annotations, and also include more video generation models in VBench to drive forward the field of video generation.
ERNIE 5.0 leads with 84.2%
Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.
Veo 3 leads with 85.1% across all independent runs
Best score per model across the official leaderboard, Epoch AI’s own runs and any board Epoch AI copied. Each bar and row names its source.
Veo 3 leads with 85.1%, claims included
Best score per model from every source, including labs’ own numbers and claims Epoch AI collected from reports and papers. Labs pick their own settings, so a claim can beat an independent run of the same model.
- 1
ERNIE 5.0 84.2%
Source: Labs’ model cards
Source: Official VBench leaderboard
Sources: Official VBench leaderboard and labs’ model cards
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | ERNIE 5.0 | 84.2% | Feb 2026 | Lab model card |
The benchmark's own leaderboard has 74 entries, 4 of them models we track. Switch the scores above to rank them together.
Number one over time 2 models have held number one across all independent runs since its release in November 2023 2 models have held number one with claims included since its release in November 2023
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official VBench leaderboard
Veo 3 on the official leaderboard passed 85% in August 2025
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Sources: Official VBench leaderboard and labs’ model cards
Labs headline it less often now, last in February 2026
Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.
| Date | Launch post | Score | As the post writes it |
|---|---|---|---|
| Feb 6, 2026 | ERNIE 5.0 | 84.2 | VBench Overall (Quality 84.40, Semantic 83.40) |
One system card reports it
Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.
| Date | Document | Score | As the document writes it | Section |
|---|---|---|---|---|
| Feb 4, 2026 | ERNIE 5.0 technical report | 84.2 | VBench Overall, Quality 84.40, Semantic 83.40 (Table 6) | Capabilities |
Versions
More in Audio and video
- Video-MMMU In use 300 expert-level lecture videos with questions testing perception, comprehension and adaptation of knowledge. Accuracy.
Qwen 3.8 Max 88.7% - LVBench In use Hours-long video understanding with multiple-choice questions on reasoning, grounding and summarization.
Gemini 4 Argon 91.7% - MMVU In use 3,000 expert-annotated questions on videos across 27 disciplines. Multiple-choice and open-ended accuracy. Qwen 3.8 Max 82.4%
- MLVU In use Multi-task long video understanding (minutes to hours) with multiple-choice and generation tasks. Accuracy. Qwen 3.8 Max 90.8%
- FLEURS In use Models transcribe read speech in 102 languages built from FLoRes-101 sentences. Model cards usually report word error rate, so lower scores are better.
Phi-4-multimodal 4.00% - VoiceBench In use Spoken instructions with speaker, environment and content variation. Voice assistant responses scored by GPT judge or rules. Qwen3.5-Omni-Plus 93.1%