Multimodal & Grounded benchmark
LVBench leaderboard
Every model the catalog carries a published LVBench value for, ranked by that value.
CategoryMultimodal & Grounded
MeasureLong-video understanding score
TasksLong-form video question answering
DifficultyExtended temporal reasoning
Published byQwen3.8-Max: A New Bar for Coding and Cowork
A long-video understanding benchmark for retrieving and reasoning over information distributed across extended video inputs.
LVBench ranking
5 models with a published LVBench 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 | Long-video understanding score |
|---|---|---|---|
| 1 |  Gemini 3.8 Flash | 87.1 | |
| 2 | Gemini 3.7 Flash | 85.4 | |
| 3 |  Qwen3.8 Max | Alibaba | 81.8 |
| 4 | Qwen3.8-Omni-Flash | Alibaba | 76.9 |
| 5 | Qwen3.8-Flash-Next | Alibaba | 76.6 |
Evidence key: Observed
Rows are ordered by the value Qwen3.8-Max: A New Bar for Coding and Cowork 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.