# Video Understanding and Reasoning on Video-MME (test) Source: [Video Understanding and Reasoning on Video-MME (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:54:26.272Z Evaluation of video understanding and reasoning capabilities on the test split of the Video-MME dataset, segmented by query length (Short, Medium, Long). 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 Understanding and Reasoning | | Dataset (including any stated split) | Video-MME (test) | | Metric | Overall Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 15 of 15 | | 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 - [Overall Accuracy](): Higher is better; unit: PERCENT. - [Accuracy (Long Query)](): Higher is better; unit: PERCENT. - [Accuracy (Medium Query)](): Higher is better; unit: PERCENT. - [Accuracy (Short Query)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Overall Accuracy | Result date | Report paper | | --- | --- | --- | --- | | SmartSight | 60.2 | Not recorded | 694a223b9eebe296e72aa952 | | Video-R1 | 59.8 | Not recorded | 694a223b9eebe296e72aa952 | | VCD | 59.8 | Not recorded | 694a223b9eebe296e72aa952 | | TCD | 59.7 | Not recorded | 694a223b9eebe296e72aa952 | | DINO-HEAL | 59.2 | Not recorded | 694a223b9eebe296e72aa952 | | SmartSight | 56.2 | Not recorded | 694a223b9eebe296e72aa952 | | Qwen2.5-VL-7B | 54.8 | Not recorded | 694a223b9eebe296e72aa952 | | VCD | 54.8 | Not recorded | 694a223b9eebe296e72aa952 | | TCD | 54.7 | Not recorded | 694a223b9eebe296e72aa952 | | DINO-HEAL | 54.6 | Not recorded | 694a223b9eebe296e72aa952 | | SmartSight | 34 | Not recorded | 694a223b9eebe296e72aa952 | | TCD | 31.1 | Not recorded | 694a223b9eebe296e72aa952 | | VCD | 30.6 | Not recorded | 694a223b9eebe296e72aa952 | | LLaVA-NEXT-Video-7B | 28.8 | Not recorded | 694a223b9eebe296e72aa952 | | DINO-HEAL | 28.3 | Not recorded | 694a223b9eebe296e72aa952 | ### SmartSight Overall Accuracy: 60.2. Result ID: 699820755698665ec47aa691. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"Video-R1"}. This record does not assert independent reproduction. ### Video-R1 Overall Accuracy: 59.8. Result ID: 699820755698665ec47aa68d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Decoding Method":"Greedy"}. This record does not assert independent reproduction. ### VCD Overall Accuracy: 59.8. Result ID: 699820755698665ec47aa690. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"Video-R1"}. This record does not assert independent reproduction. ### TCD Overall Accuracy: 59.7. Result ID: 699820755698665ec47aa68f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"Video-R1"}. This record does not assert independent reproduction. ### DINO-HEAL Overall Accuracy: 59.2. Result ID: 699820755698665ec47aa68e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"Video-R1"}. This record does not assert independent reproduction. ### SmartSight Overall Accuracy: 56.2. Result ID: 699820755698665ec47aa68c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-7B"}. This record does not assert independent reproduction. ### Qwen2.5-VL-7B Overall Accuracy: 54.8. Result ID: 699820755698665ec47aa688. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Decoding Method":"Greedy"}. This record does not assert independent reproduction. ### VCD Overall Accuracy: 54.8. Result ID: 699820755698665ec47aa68b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-7B"}. This record does not assert independent reproduction. ### TCD Overall Accuracy: 54.7. Result ID: 699820755698665ec47aa68a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-7B"}. This record does not assert independent reproduction. ### DINO-HEAL Overall Accuracy: 54.6. Result ID: 699820755698665ec47aa689. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-7B"}. This record does not assert independent reproduction. ### SmartSight Overall Accuracy: 34. Result ID: 699820755698665ec47aa696. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"LLaVA-NEXT-Video-7B"}. This record does not assert independent reproduction. ### TCD Overall Accuracy: 31.1. Result ID: 699820755698665ec47aa694. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"LLaVA-NEXT-Video-7B"}. This record does not assert independent reproduction. ### VCD Overall Accuracy: 30.6. Result ID: 699820755698665ec47aa695. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"LLaVA-NEXT-Video-7B"}. This record does not assert independent reproduction. ### LLaVA-NEXT-Video-7B Overall Accuracy: 28.8. Result ID: 699820755698665ec47aa692. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Decoding Method":"Greedy"}. This record does not assert independent reproduction. ### DINO-HEAL Overall Accuracy: 28.3. Result ID: 699820755698665ec47aa693. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"LLaVA-NEXT-Video-7B"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.