Multimodal & Grounded benchmark
VideoMMMU leaderboard
Every model the catalog carries a published VideoMMMU value for, ranked by that value.
CategoryMultimodal & Grounded
MeasureVideo + text reasoning
TasksVideo-grounded expert reasoning
DifficultyFrontier multimodal video reasoning
Published byQwen3.6 launch benchmarks
A video extension of MMMU-style multimodal reasoning over expert questions grounded in temporal media.
VideoMMMU ranking
11 models with a published VideoMMMU value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.
| Rank | Model | Provider | Video + text reasoning |
|---|---|---|---|
| 1 |  Qwen3.8 Max | Alibaba | 88.7 |
| 2 |  Gemini 3 Pro | 87.6 | |
| 3 | DSdots3-note Preview | Dots Studio | 86.8 |
| 4 |  Kimi K2.5 | Moonshot AI | 86.6 |
| 5 | Qwen3.7 Plus | Alibaba | 85.4 |
| 6 | Qwen3.5 397B | Alibaba | 84.7 |
| 7 |  MiniMax M3 | MiniMax | 84.6 |
| 8 |  Claude Opus 4.5 | Anthropic | 84.4 |
| 8 | Qwen3.6-27B | Alibaba | 84.4 |
| 10 | Qwen3.6 Plus | Alibaba | 84.0 |
| 11 | Qwen3.6-35B-A3B | Alibaba | 83.7 |
Evidence key: Observed
Rows are ordered by the value Qwen3.6 launch benchmarks published, highest first. The source does not state whether a higher value is the better result, so this page does not either: for a benchmark that measures a rate of failure, read the table from the bottom.