{"benchmark_id":"mmbench-video","name":"MMBench-Video","parent_benchmark":null,"categories":["multimodal","reasoning","video","vision"],"modality":"multimodal","multilingual":false,"max_score":1.0,"language":"en","description":"A long-form multi-shot benchmark for holistic video understanding that incorporates approximately 600 web videos from YouTube spanning 16 major categories, with each video ranging from 30 seconds to 6 minutes. Includes roughly 2,000 original question-answer pairs covering 26 fine-grained capabilities.","paper_link":"https://arxiv.org/abs/2406.14515","implementation_link":null,"verified":false,"created_at":"2026-05-07T16:53:24.651449+00:00","updated_at":"2026-07-05T18:27:54.066899+00:00","statistics":{"total_models":3,"average_score":0.019133333333333332,"min_score":0.0179,"max_score":0.0202,"score_stddev":0.0011590225767142475,"verified_count":0,"self_reported_count":3},"child_benchmarks":[],"linked_dataset":null,"models":[{"rank":1,"model_id":"qwen2.5-vl-72b","model_name":"Qwen2.5 VL 72B Instruct","organization_id":"qwen","organization_name":"Alibaba Cloud / Qwen Team","organization_country":"CN","score":0.0202,"normalized_score":0.0202,"verified":false,"self_reported":true,"self_reported_source":"https://github.com/QwenLM/Qwen2.5-VL","analysis_method":"Score","verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2025-01-26","param_count":72000000000,"is_open_source":true,"is_new":false,"best_latency":null,"latency_provider":null,"best_throughput":null,"throughput_provider":null,"context_provider":null},{"rank":2,"model_id":"qwen2.5-vl-32b","model_name":"Qwen2.5 VL 32B Instruct","organization_id":"qwen","organization_name":"Alibaba Cloud / Qwen Team","organization_country":"CN","score":0.0193,"normalized_score":0.0193,"verified":false,"self_reported":true,"self_reported_source":"https://github.com/QwenLM/Qwen2.5-VL","analysis_method":"Score","verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2025-02-28","param_count":33500000000,"is_open_source":true,"is_new":false,"best_latency":null,"latency_provider":null,"best_throughput":null,"throughput_provider":null,"context_provider":null},{"rank":3,"model_id":"qwen2.5-vl-7b","model_name":"Qwen2.5 VL 7B Instruct","organization_id":"qwen","organization_name":"Alibaba Cloud / Qwen Team","organization_country":"CN","score":0.0179,"normalized_score":0.0179,"verified":false,"self_reported":true,"self_reported_source":"https://github.com/QwenLM/Qwen2.5-VL","analysis_method":"Score","verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2025-01-26","param_count":8290000000,"is_open_source":true,"is_new":false,"best_latency":null,"latency_provider":null,"best_throughput":null,"throughput_provider":null,"context_provider":null}]}