VideoQA on LongVideoBench (Medium/Long/All Scores)
72.6Score (All Lengths)GPT-5
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GPT-5LLM=-2026.03 | 72.6 | — | — | |
| GPT-4oLLM=-2026.03 | 66.7 | 69.1 | 60.9 | |
| Gemini-1.5-ProLLM=-2026.03 | 64 | 65.3 | 58.6 | |
| TSPOLLM=7B2026.03 | 63.9 | — | — | |
| LLaVA-Video + FlexMemLLM=7B2026.03 | 63.6 | 65.5 | 57.3 | |
| LLaVA-VideoLLM=7B2026.03 | 60 | 60.7 | 52.1 | |
| TimeMarkerLLM=8B2026.03 | 56.3 | — | — | |
| Qwen2.5-VLLLM=7B2026.03 | 56 | — | — | |
| mPLUG-Owl3LLM=7B2026.03 | 52.1 | — | — | |
| Video-XLLLM=7B2026.03 | 50.5 | 49 | 45.2 |