筑梦

中|En

Vchitect 2.0Embark on a Visual Fantasy Journey20-second video generation, flexible aspect ratios, generative space-time enhancement, long video evaluationTechnical ReportGitHub CodeTraining Framework Try It

Text-to-Video

Vchitect2.0, developed by the Shanghai AI Lab, is an advanced video generation model designed to empower video creation.

5 to 20 seconds video generation

Flexible aspect ratios, allowing users to generate videos of arbitrary aspect ratios.

High-definition quality and integrated super-resolution and frame insertion, featuring user-adjustable content correction capabilities.

Image-to-Video

Transform static images into 5-10 second videos. It enables users to effortlessly convert photos or designs into captivating visual experiences.

Video Generation Benchmark -- VBench

VBench now supports evaluating long video generative models.

Comprehensive and Continuously Updated Evaluation Leaderboard

Covering 28 text-to-video models and 12 image-to-video models.

Open Source, One-click Evaluation Deployment

Recognized as the industry standard for automated video generation evaluation and extensively reported by media outlets like the South China Morning Post. The evaluation suite has been adopted by major video generation models, enhancing evaluation consistency and transparency for video generation.

Supports Multiple Long Video Models

VBench upgraded its evaluation suite to support mainstream long video generation models, now including support for models like Gen-3, Kling, OpenSora, etc.

Model	Total Score	Quality Score	Semantic Score	Subject Consistency	Background Consistency	Temporal Flickering	Motion Smoothness	Dynamic Degree	Aesthetic Quality	Imaging Quality	Object Class	Multiple Objects	Human Action	Color	Spatial Relationship	Scene	Appearance Style	Temporal Style	Overall Consistency

Gen-3	82.32%	84.11%	75.17%	97.10%	96.62%	98.61%	99.23%	60.14%	63.34%	66.82%	87.81%	53.64%	96.40%	80.90%	65.09%	54.57%	24.31%	24.71%	26.69%
Kling (2024-07 high-performance mode)	81.85%	83.39%	75.68%	98.33%	97.60%	99.30%	99.40%	46.94%	61.21%	65.62%	87.24%	68.05%	93.40%	89.90%	73.03%	50.86%	19.62%	24.17%	26.42%
CogVideoX-5B-SAT (SAT prompt-optimized)	81.61%	82.75%	77.04%	96.23%	96.52%	98.66%	96.92%	70.97%	61.98%	62.90%	85.23%	62.11%	99.40%	82.81%	66.35%	53.20%	24.91%	25.38%	27.59%
Vchitect 2.0-2B	81.57%	82.51%	77.79%	96.42%	96.53%	98.45%	97.76%	58.33%	61.47%	65.60%	87.81%	69.35%	97.00%	86.87%	54.64%	57.51%	24.93%	25.56%	28.01%
CogVideoX-2B-SAT (SAT prompt-optimized)	80.91%	82.18%	75.83%	96.78%	96.63%	98.89%	97.73%	59.86%	60.82%	61.68%	83.37%	62.63%	98.00%	79.41%	69.90%	51.14%	24.80%	24.36%	26.66%
OpenSorav1-2 (8s)	79.76%	81.35%	73.39%	96.75%	97.61%	99.53%	98.50%	42.39%	56.85%	63.34%	82.22%	51.83%	91.20%	90.08%	68.56%	42.44%	23.95%	24.54%	26.85%
OpenSoraPlanv1-1	78.00%	80.91%	66.38%	95.73%	96.73%	99.03%	98.28%	47.72%	56.85%	62.28%	76.30%	40.35%	86.80%	89.19%	53.11%	27.17%	22.90%	23.87%	26.52%
OpenSorav1-1	75.66%	77.75%	67.35%	92.35%	97.52%	98.31%	92.78%	82.14%	50.12%	54.90%	86.76%	40.97%	84.20%	74.56%	52.47%	38.63%	23.50%	23.86%	26.37%
Mira384	71.87%	78.83%	44.21%	96.23%	96.92%	98.29%	97.54%	60.33%	42.51%	60.16%	52.06%	12.52%	63.80%	42.24%	27.83%	16.34%	21.89%	18.77%	18.72%

vchitect@pjlab.org.cn