# Long Video Understanding on LVBench (val)

Source: [Long Video Understanding on LVBench (val)](<https://www.sota2.com/research/sota/long-video-understanding-on-lvbench-val>)

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

Data updated: 2026-02-26T06:03:48.693Z

Evaluation of Long Video Understanding capabilities using the validation split of the LVBench dataset, measured by Score.

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 | Long Video Understanding |
| Dataset (including any stated split) | LVBench (val) |
| Metric | Score |
| 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

- [Score](<https://www.sota2.com/research/sota/long-video-understanding-on-lvbench-val.md?metric=Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Score | Result date | Report paper |
| --- | --- | --- | --- |
| GLM-4V-Plus | 58.7 | Not recorded | 68fa793d1e209131cea7ccf7 |
| Qwen2.5-VL+AdaRETAKE | 53.3 | Not recorded | 68fa793d1e209131cea7ccf7 |
| Qwen2.5-VL+AdaRETAKE | 51.2 | Not recorded | 68fa793d1e209131cea7ccf7 |
| LLaVA-Video+AdaRETAKE | 49.6 | Not recorded | 68fa793d1e209131cea7ccf7 |
| Qwen2-VL+AdaRETAKE | 48.9 | Not recorded | 68fa793d1e209131cea7ccf7 |
| Qwen2.5-VL | 47.3 | Not recorded | 68fa793d1e209131cea7ccf7 |
| VideoLLaMA3 | 45.3 | Not recorded | 68fa793d1e209131cea7ccf7 |
| Qwen2.5-VL | 45.3 | Not recorded | 68fa793d1e209131cea7ccf7 |
| InternVL2.5 | 43.6 | Not recorded | 68fa793d1e209131cea7ccf7 |
| LLaVA-Video | 43.1 | Not recorded | 68fa793d1e209131cea7ccf7 |
| Qwen2-VL | 42.4 | Not recorded | 68fa793d1e209131cea7ccf7 |
| Qwen2-VL | 41.3 | Not recorded | 68fa793d1e209131cea7ccf7 |
| Gemini-1.5-Pro | 33.1 | Not recorded | 68fa793d1e209131cea7ccf7 |
| Oryx-1.5 | 30.4 | Not recorded | 68fa793d1e209131cea7ccf7 |
| GPT-4o | 27 | Not recorded | 68fa793d1e209131cea7ccf7 |

### GLM-4V-Plus

Score: 58.7. Result ID: 6995d5173badfaffe6dab642.

[Reporting paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Qwen2.5-VL+AdaRETAKE

Score: 53.3. Result ID: 6995d5173badfaffe6dab650.

[Reporting paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

[Implementation](<https://github.com/SCZwangxiao/video-FlexReduc.git>)

Recorded evaluation configuration: {"LLM Size":"72B"}. This record does not assert independent reproduction.

### Qwen2.5-VL+AdaRETAKE

Score: 51.2. Result ID: 6995d5173badfaffe6dab64b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

[Implementation](<https://github.com/SCZwangxiao/video-FlexReduc.git>)

Recorded evaluation configuration: {"LLM Size":"7B"}. This record does not assert independent reproduction.

### LLaVA-Video+AdaRETAKE

Score: 49.6. Result ID: 6995d5173badfaffe6dab647.

[Reporting paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

[Implementation](<https://github.com/SCZwangxiao/video-FlexReduc.git>)

Recorded evaluation configuration: {"LLM Size":"7B"}. This record does not assert independent reproduction.

### Qwen2-VL+AdaRETAKE

Score: 48.9. Result ID: 6995d5173badfaffe6dab649.

[Reporting paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

[Implementation](<https://github.com/SCZwangxiao/video-FlexReduc.git>)

Recorded evaluation configuration: {"LLM Size":"7B"}. This record does not assert independent reproduction.

### Qwen2.5-VL

Score: 47.3. Result ID: 6995d5173badfaffe6dab64f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

Recorded evaluation configuration: {"LLM Size":"72B"}. This record does not assert independent reproduction.

### VideoLLaMA3

Score: 45.3. Result ID: 6995d5173badfaffe6dab645.

[Reporting paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

Recorded evaluation configuration: {"LLM Size":"7B"}. This record does not assert independent reproduction.

### Qwen2.5-VL

Score: 45.3. Result ID: 6995d5173badfaffe6dab64a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

Recorded evaluation configuration: {"LLM Size":"7B"}. This record does not assert independent reproduction.

### InternVL2.5

Score: 43.6. Result ID: 6995d5173badfaffe6dab64e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

Recorded evaluation configuration: {"LLM Size":"72B"}. This record does not assert independent reproduction.

### LLaVA-Video

Score: 43.1. Result ID: 6995d5173badfaffe6dab646.

[Reporting paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a561e209131cea82934.md>)

[Implementation](<https://llava-vl.github.io/blog/llava-video>)

Recorded evaluation configuration: {"LLM Size":"7B"}. This record does not assert independent reproduction.

### Qwen2-VL

Score: 42.4. Result ID: 6995d5173badfaffe6dab648.

[Reporting paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75151e209131cea65976.md>)

[Implementation](<https://github.com/QwenLM/Qwen2-VL>)

Recorded evaluation configuration: {"LLM Size":"7B"}. This record does not assert independent reproduction.

### Qwen2-VL

Score: 41.3. Result ID: 6995d5173badfaffe6dab64d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75151e209131cea65976.md>)

[Implementation](<https://github.com/QwenLM/Qwen2-VL>)

Recorded evaluation configuration: {"LLM Size":"72B"}. This record does not assert independent reproduction.

### Gemini-1.5-Pro

Score: 33.1. Result ID: 6995d5173badfaffe6dab644.

[Reporting paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Oryx-1.5

Score: 30.4. Result ID: 6995d5173badfaffe6dab64c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

Recorded evaluation configuration: {"LLM Size":"32B"}. This record does not assert independent reproduction.

### GPT-4o

Score: 27. Result ID: 6995d5173badfaffe6dab643.

[Reporting paper](<https://www.sota2.com/research/paper/68fa793d1e209131cea7ccf7.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

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