{"record": {"aliases": [], "artifacts": [], "categories": ["multimodal", "reasoning", "spatial_reasoning", "video", "vision"], "description": {"en": "A comprehensive multi-modal video understanding benchmark covering 20 challenging video tasks that require temporal understanding beyond single-frame analysis. Tasks span from perception to cognition, including action recognition, temporal reasoning, spatial reasoning, object interaction, scene transition, and counterfactual inference. Uses a novel static-to-dynamic method to systematically generate video tasks from existing annotations."}, "documents": [{"document_type": "registry_page", "id": "llm_stats:https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "source": "llm_stats", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "title": "MVBench"}], "evidence_summary": {"document_count": 1, "model_count": 17, "model_count_basis": "source_model_id"}, "key": "llm-stats:mvbench", "modality": "multimodal", "name": "MVBench", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"crawl_bundle": "llm_stats_2026-08-17", "crawled_at": "2026-08-17T18:49:35.489925+00:00", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500"}, "publisher": null, "released": "2023-11-28", "released_reference": {"basis": "paper_first_version", "note": "First version of the introducing paper: MVBench: A Comprehensive Multi-modal Video Understanding Benchmark. The source-record description identifies this benchmark or its original partition; no dates or scores are joined by name at runtime.", "source_key": "llm-stats:mvbench", "source_url": "https://arxiv.org/abs/2311.17005"}, "schema_version": 1, "sizes": [], "slug": "llm-stats-mvbench", "source": "llm_stats", "source_benchmark_id": "mvbench"}, "schema_version": 1, "scores_by_source": {"llm_stats": {"rows": [{"comparable_group": null, "crawled_at": "2026-08-17T18:49:35.489925+00:00", "date_precision": "model_announcement", "document_id": "llm_stats:https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "key": "llm-stats:mvbench", "measured_by": null, "model_id": "qwen3.5-122b-a10b", "model_name": "Qwen3.5-122B-A10B", "obs_id": "llm_stats:mvbench:qwen3.5-122b-a10b", "organization": "Qwen", "rank_in_source_response": 1, "raw_value": "0.766", "reported_by": "self_reported", "reported_date": "2026-02-24", "series_id": "llm_stats:mvbench:default", "source": "llm_stats", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "value": 0.766, "value_kind": "number"}, {"comparable_group": null, "crawled_at": "2026-08-17T18:49:35.489925+00:00", "date_precision": "model_announcement", "document_id": "llm_stats:https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "key": "llm-stats:mvbench", "measured_by": null, "model_id": "qwen3.6-27b", "model_name": "Qwen3.6-27B", "obs_id": "llm_stats:mvbench:qwen3.6-27b", "organization": "Qwen", "rank_in_source_response": 2, "raw_value": "0.755", "reported_by": "self_reported", "reported_date": "2026-04-21", "series_id": "llm_stats:mvbench:default", "source": "llm_stats", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "value": 0.755, "value_kind": "number"}, {"comparable_group": null, "crawled_at": "2026-08-17T18:49:35.489925+00:00", "date_precision": "model_announcement", "document_id": "llm_stats:https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "key": "llm-stats:mvbench", "measured_by": null, "model_id": "qwen3.5-35b-a3b", "model_name": "Qwen3.5-35B-A3B", "obs_id": "llm_stats:mvbench:qwen3.5-35b-a3b", "organization": "Qwen", "rank_in_source_response": 3, "raw_value": "0.748", "reported_by": "self_reported", "reported_date": "2026-02-24", "series_id": "llm_stats:mvbench:default", "source": "llm_stats", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "value": 0.748, "value_kind": "number"}, {"comparable_group": null, "crawled_at": "2026-08-17T18:49:35.489925+00:00", "date_precision": "model_announcement", "document_id": "llm_stats:https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "key": "llm-stats:mvbench", "measured_by": null, "model_id": "qwen3.6-35b-a3b", "model_name": "Qwen3.6-35B-A3B", "obs_id": "llm_stats:mvbench:qwen3.6-35b-a3b", "organization": "Qwen", "rank_in_source_response": 4, "raw_value": "0.746", "reported_by": "self_reported", "reported_date": "2026-04-16", "series_id": "llm_stats:mvbench:default", "source": "llm_stats", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "value": 0.746, "value_kind": "number"}, {"comparable_group": null, "crawled_at": "2026-08-17T18:49:35.489925+00:00", "date_precision": "model_announcement", "document_id": "llm_stats:https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "key": "llm-stats:mvbench", "measured_by": null, "model_id": "qwen3.5-27b", "model_name": "Qwen3.5-27B", "obs_id": "llm_stats:mvbench:qwen3.5-27b", "organization": "Qwen", "rank_in_source_response": 5, "raw_value": "0.746", "reported_by": "self_reported", "reported_date": "2026-02-24", "series_id": "llm_stats:mvbench:default", "source": "llm_stats", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "value": 0.746, "value_kind": "number"}, {"comparable_group": null, "crawled_at": "2026-08-17T18:49:35.489925+00:00", "date_precision": "model_announcement", "document_id": "llm_stats:https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "key": "llm-stats:mvbench", "measured_by": null, "model_id": "qwen2-vl-72b", "model_name": "Qwen2-VL-72B-Instruct", "obs_id": "llm_stats:mvbench:qwen2-vl-72b", "organization": "Qwen", "rank_in_source_response": 6, "raw_value": "0.736", "reported_by": "self_reported", "reported_date": "2024-08-29", "series_id": "llm_stats:mvbench:default", "source": "llm_stats", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "value": 0.736, "value_kind": "number"}, {"comparable_group": null, "crawled_at": "2026-08-17T18:49:35.489925+00:00", "date_precision": "model_announcement", "document_id": "llm_stats:https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "key": "llm-stats:mvbench", "measured_by": null, "model_id": "qwen3-vl-32b-thinking", "model_name": "Qwen3 VL 32B Thinking", "obs_id": "llm_stats:mvbench:qwen3-vl-32b-thinking", "organization": "Qwen", "rank_in_source_response": 7, "raw_value": "0.732", "reported_by": "self_reported", "reported_date": "2025-09-22", "series_id": "llm_stats:mvbench:default", "source": "llm_stats", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "value": 0.732, "value_kind": "number"}, {"comparable_group": null, "crawled_at": "2026-08-17T18:49:35.489925+00:00", "date_precision": "model_announcement", "document_id": "llm_stats:https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "key": "llm-stats:mvbench", "measured_by": null, "model_id": "qwen3-vl-32b-instruct", "model_name": "Qwen3 VL 32B Instruct", "obs_id": "llm_stats:mvbench:qwen3-vl-32b-instruct", "organization": "Qwen", "rank_in_source_response": 8, "raw_value": "0.728", "reported_by": "self_reported", "reported_date": "2025-09-22", "series_id": "llm_stats:mvbench:default", "source": "llm_stats", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "value": 0.728, "value_kind": "number"}, {"comparable_group": null, "crawled_at": "2026-08-17T18:49:35.489925+00:00", "date_precision": "model_announcement", "document_id": "llm_stats:https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "key": "llm-stats:mvbench", "measured_by": null, "model_id": "qwen3-vl-30b-a3b-instruct", "model_name": "Qwen3 VL 30B A3B Instruct", "obs_id": "llm_stats:mvbench:qwen3-vl-30b-a3b-instruct", "organization": "Qwen", "rank_in_source_response": 9, "raw_value": "0.723", "reported_by": "self_reported", "reported_date": "2025-09-22", "series_id": "llm_stats:mvbench:default", "source": "llm_stats", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "value": 0.723, "value_kind": "number"}, {"comparable_group": null, "crawled_at": "2026-08-17T18:49:35.489925+00:00", "date_precision": "model_announcement", "document_id": "llm_stats:https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "key": "llm-stats:mvbench", "measured_by": null, "model_id": "qwen3-vl-30b-a3b-thinking", "model_name": "Qwen3 VL 30B A3B Thinking", "obs_id": "llm_stats:mvbench:qwen3-vl-30b-a3b-thinking", "organization": "Qwen", "rank_in_source_response": 10, "raw_value": "0.72", "reported_by": "self_reported", "reported_date": "2025-09-22", "series_id": "llm_stats:mvbench:default", "source": "llm_stats", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "value": 0.72, "value_kind": "number"}, {"comparable_group": null, "crawled_at": "2026-08-17T18:49:35.489925+00:00", "date_precision": "model_announcement", "document_id": "llm_stats:https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "key": "llm-stats:mvbench", "measured_by": null, "model_id": "qwen2.5-vl-72b", "model_name": "Qwen2.5 VL 72B Instruct", "obs_id": "llm_stats:mvbench:qwen2.5-vl-72b", "organization": "Qwen", "rank_in_source_response": 11, "raw_value": "0.704", "reported_by": "self_reported", "reported_date": "2025-01-26", "series_id": "llm_stats:mvbench:default", "source": "llm_stats", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "value": 0.704, "value_kind": "number"}, {"comparable_group": null, "crawled_at": "2026-08-17T18:49:35.489925+00:00", "date_precision": "model_announcement", "document_id": "llm_stats:https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "key": "llm-stats:mvbench", "measured_by": null, "model_id": "qwen2.5-omni-7b", "model_name": "Qwen2.5-Omni-7B", "obs_id": "llm_stats:mvbench:qwen2.5-omni-7b", "organization": "Qwen", "rank_in_source_response": 12, "raw_value": "0.703", "reported_by": "self_reported", "reported_date": "2025-03-27", "series_id": "llm_stats:mvbench:default", "source": "llm_stats", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "value": 0.703, "value_kind": "number"}, {"comparable_group": null, "crawled_at": "2026-08-17T18:49:35.489925+00:00", "date_precision": "model_announcement", "document_id": "llm_stats:https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "key": "llm-stats:mvbench", "measured_by": null, "model_id": "qwen2.5-vl-7b", "model_name": "Qwen2.5 VL 7B Instruct", "obs_id": "llm_stats:mvbench:qwen2.5-vl-7b", "organization": "Qwen", "rank_in_source_response": 13, "raw_value": "0.696", "reported_by": "self_reported", "reported_date": "2025-01-26", "series_id": "llm_stats:mvbench:default", "source": "llm_stats", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "value": 0.696, "value_kind": "number"}, {"comparable_group": null, "crawled_at": "2026-08-17T18:49:35.489925+00:00", "date_precision": "model_announcement", "document_id": "llm_stats:https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "key": "llm-stats:mvbench", "measured_by": null, "model_id": "qwen3-vl-4b-thinking", "model_name": "Qwen3 VL 4B Thinking", "obs_id": "llm_stats:mvbench:qwen3-vl-4b-thinking", "organization": "Qwen", "rank_in_source_response": 14, "raw_value": "0.693", "reported_by": "self_reported", "reported_date": "2025-09-22", "series_id": "llm_stats:mvbench:default", "source": "llm_stats", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "value": 0.693, "value_kind": "number"}, {"comparable_group": null, "crawled_at": "2026-08-17T18:49:35.489925+00:00", "date_precision": "model_announcement", "document_id": "llm_stats:https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "key": "llm-stats:mvbench", "measured_by": null, "model_id": "qwen3-vl-8b-thinking", "model_name": "Qwen3 VL 8B Thinking", "obs_id": "llm_stats:mvbench:qwen3-vl-8b-thinking", "organization": "Qwen", "rank_in_source_response": 15, "raw_value": "0.69", "reported_by": "self_reported", "reported_date": "2025-09-22", "series_id": "llm_stats:mvbench:default", "source": "llm_stats", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "value": 0.69, "value_kind": "number"}, {"comparable_group": null, "crawled_at": "2026-08-17T18:49:35.489925+00:00", "date_precision": "model_announcement", "document_id": "llm_stats:https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "key": "llm-stats:mvbench", "measured_by": null, "model_id": "qwen3-vl-4b-instruct", "model_name": "Qwen3 VL 4B Instruct", "obs_id": "llm_stats:mvbench:qwen3-vl-4b-instruct", "organization": "Qwen", "rank_in_source_response": 16, "raw_value": "0.689", "reported_by": "self_reported", "reported_date": "2025-09-22", "series_id": "llm_stats:mvbench:default", "source": "llm_stats", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "value": 0.689, "value_kind": "number"}, {"comparable_group": null, "crawled_at": "2026-08-17T18:49:35.489925+00:00", "date_precision": "model_announcement", "document_id": "llm_stats:https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "key": "llm-stats:mvbench", "measured_by": null, "model_id": "qwen3-vl-8b-instruct", "model_name": "Qwen3 VL 8B Instruct", "obs_id": "llm_stats:mvbench:qwen3-vl-8b-instruct", "organization": "Qwen", "rank_in_source_response": 17, "raw_value": "0.687", "reported_by": "self_reported", "reported_date": "2025-09-22", "series_id": "llm_stats:mvbench:default", "source": "llm_stats", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500", "value": 0.687, "value_kind": "number"}], "series": {"bounds": {"basis": "aggregator_declared", "max": 1.0, "min": null}, "declared_max": 1.0, "direction": "higher_is_better", "direction_basis": "source_rank_descending", "display_scale": null, "first_score_record": {"date_precision": "model_announcement", "obs_id": "llm_stats:mvbench:qwen2-vl-72b", "reported_at": "2024-08-29", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500"}, "first_score_report": null, "key": "llm-stats:mvbench", "max_score_contradicted": false, "metric": null, "observation_count": 17, "observed_max": 0.766, "score_summary": {"display_max": 76.6, "display_multiplier": 100, "model_count": 17, "model_count_basis": "source_model_id", "numeric_count": 17, "raw_max": 0.766, "raw_min": 0.687, "source_reference": {"crawled_at": "2026-08-17T18:49:35.489925+00:00", "obs_id": "llm_stats:mvbench:qwen3.5-122b-a10b", "reported_date": "2026-02-24", "source_url": "https://api.zeroeval.com/leaderboard/benchmarks/mvbench?top_n=500"}, "unit": null}, "series_id": "llm_stats:mvbench:default"}}}, "siblings": []}
