# Video Understanding on VideoMME w/ sub (Accuracy)

Source: [Video Understanding on VideoMME w/ sub (Accuracy)](<https://www.sota2.com/research/sota/video-understanding-on-videomme-w-sub-accuracy>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-05-21T10:44:48.602Z

Evaluation of video understanding capabilities on the VideoMME dataset subset including subtitles, measured specifically by Accuracy.

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 |
| Dataset (including any stated split) | VideoMME w/ sub |
| Metric | 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

- [Accuracy](<https://www.sota2.com/research/sota/video-understanding-on-videomme-w-sub-accuracy.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Gemini 1.5 Pro | 81.3 | Not recorded | 6a0e9193f984739ed6738ad0 |
| GPT-4o | 77.2 | Not recorded | 6a0e9193f984739ed6738ad0 |
| ParaVT-8B | 69.4 | Not recorded | 6a0e9193f984739ed6738ad0 |
| Qwen3-VL-8B | 68.4 | Not recorded | 6a0e9193f984739ed6738ad0 |
| Time-R1-7B | 66.2 | Not recorded | 6a0e9193f984739ed6738ad0 |
| Video-R1-7B | 66 | Not recorded | 6a0e9193f984739ed6738ad0 |
| LongVT-RFT-7B | 66 | Not recorded | 6a0e9193f984739ed6738ad0 |
| VideoRFT-7B | 65.6 | Not recorded | 6a0e9193f984739ed6738ad0 |
| Video-Thinker-7B | 65.3 | Not recorded | 6a0e9193f984739ed6738ad0 |
| ReWatch-R1-7B | 65 | Not recorded | 6a0e9193f984739ed6738ad0 |
| Qwen2.5-VL-7B | 64.5 | Not recorded | 6a0e9193f984739ed6738ad0 |
| Conan-7B | 62.8 | Not recorded | 6a0e9193f984739ed6738ad0 |
| VideoChat-R1-7B | 58.2 | Not recorded | 6a0e9193f984739ed6738ad0 |
| SAGE-7B | 52.4 | Not recorded | 6a0e9193f984739ed6738ad0 |
| VideoZoomer-7B | 48.3 | Not recorded | 6a0e9193f984739ed6738ad0 |

### Gemini 1.5 Pro

Accuracy: 81.3. Result ID: 6a0ee15efbc373592c06d935.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e9193f984739ed6738ad0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76361e209131cea6be11.md>)

Recorded evaluation configuration: {"Setting":"best-setting numbers from official reports"}. This record does not assert independent reproduction.

### GPT-4o

Accuracy: 77.2. Result ID: 6a0ee15efbc373592c06d934.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e9193f984739ed6738ad0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75681e209131cea67a98.md>)

Recorded evaluation configuration: {"Setting":"best-setting numbers from official reports"}. This record does not assert independent reproduction.

### ParaVT-8B

Accuracy: 69.4. Result ID: 6a0ee15efbc373592c06d942.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e9193f984739ed6738ad0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0e9193f984739ed6738ad0.md>)

[Implementation](<https://github.com/EvolvingLMMs-Lab/ParaVT>)

Recorded evaluation configuration: {"Setting":"tool-augmented setting (\<think\>→\<tool\_call\>→\<answer\>)"}. This record does not assert independent reproduction.

### Qwen3-VL-8B

Accuracy: 68.4. Result ID: 6a0ee15efbc373592c06d93d.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e9193f984739ed6738ad0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6927db54d6b0dd3342930f66.md>)

Recorded evaluation configuration: {"Setting":"tool-augmented setting (\<think\>→\<tool\_call\>→\<answer\>)"}. This record does not assert independent reproduction.

### Time-R1-7B

Accuracy: 66.2. Result ID: 6a0ee15efbc373592c06d93a.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e9193f984739ed6738ad0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79b81e209131cea7f1a6.md>)

Recorded evaluation configuration: {"Setting":"reasoning-enhanced setting (\<think\>→\<answer\>)"}. This record does not assert independent reproduction.

### Video-R1-7B

Accuracy: 66. Result ID: 6a0ee15efbc373592c06d937.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e9193f984739ed6738ad0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c6f1e209131cea8e0ec.md>)

Recorded evaluation configuration: {"Setting":"reasoning-enhanced setting (\<think\>→\<answer\>)"}. This record does not assert independent reproduction.

### LongVT-RFT-7B

Accuracy: 66. Result ID: 6a0ee15efbc373592c06d93f.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e9193f984739ed6738ad0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6927db35d6b0dd3342930e0c.md>)

[Implementation](<https://github.com/EvolvingLMMs-Lab/LongVT>)

Recorded evaluation configuration: {"Setting":"tool-augmented setting (\<think\>→\<tool\_call\>→\<answer\>)"}. This record does not assert independent reproduction.

### VideoRFT-7B

Accuracy: 65.6. Result ID: 6a0ee15efbc373592c06d939.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e9193f984739ed6738ad0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c241e209131cea8ca8e.md>)

Recorded evaluation configuration: {"Setting":"reasoning-enhanced setting (\<think\>→\<answer\>)"}. This record does not assert independent reproduction.

### Video-Thinker-7B

Accuracy: 65.3. Result ID: 6a0ee15efbc373592c06d93c.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e9193f984739ed6738ad0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69004e133ef7ccd74e5e2e5d.md>)

Recorded evaluation configuration: {"Setting":"reasoning-enhanced setting (\<think\>→\<answer\>)"}. This record does not assert independent reproduction.

### ReWatch-R1-7B

Accuracy: 65. Result ID: 6a0ee15efbc373592c06d93b.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e9193f984739ed6738ad0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bc11e209131cea8a357.md>)

Recorded evaluation configuration: {"Setting":"reasoning-enhanced setting (\<think\>→\<answer\>)"}. This record does not assert independent reproduction.

### Qwen2.5-VL-7B

Accuracy: 64.5. Result ID: 6a0ee15efbc373592c06d936.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e9193f984739ed6738ad0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa773a1e209131cea71d3c.md>)

Recorded evaluation configuration: {"Setting":"direct-answer setting (native single-pass prompt)"}. This record does not assert independent reproduction.

### Conan-7B

Accuracy: 62.8. Result ID: 6a0ee15efbc373592c06d93e.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e9193f984739ed6738ad0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fc596c3ef7ccd74e45262b.md>)

Recorded evaluation configuration: {"Setting":"tool-augmented setting (\<think\>→\<tool\_call\>→\<answer\>)"}. This record does not assert independent reproduction.

### VideoChat-R1-7B

Accuracy: 58.2. Result ID: 6a0ee15efbc373592c06d938.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e9193f984739ed6738ad0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7ba11e209131cea89767.md>)

Recorded evaluation configuration: {"Setting":"reasoning-enhanced setting (\<think\>→\<answer\>)"}. This record does not assert independent reproduction.

### SAGE-7B

Accuracy: 52.4. Result ID: 6a0ee15efbc373592c06d940.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e9193f984739ed6738ad0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6942391b456146936285064b.md>)

[Implementation](<https://github.com/allenai/SAGE>)

Recorded evaluation configuration: {"Setting":"tool-augmented setting (\<think\>→\<tool\_call\>→\<answer\>)"}. This record does not assert independent reproduction.

### VideoZoomer-7B

Accuracy: 48.3. Result ID: 6a0ee15efbc373592c06d941.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e9193f984739ed6738ad0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69535c7e53fdd5c36173d88e.md>)

Recorded evaluation configuration: {"Setting":"tool-augmented setting (\<think\>→\<tool\_call\>→\<answer\>)"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.
