Inference
EN

MLVU mean average

A multi-task video understanding benchmark averaged across MLVU categories.

ModelsModels with a result.10
Top resultBest result on this benchmark.—
Top-3 spreadPoints from first to third.—
YearYear of release.—

Result and price

8 of 10 results

Results

10 results
#ModelResultThe score from the source.
—Qwen3.8 Max90.8%
—Qwen3.7 Plus87.4%
—Qwen3.5 397B86.7%
—Qwen3.6-27B86.6%
—Qwen3.6-35B-A3B86.2%
—GPT-5.285.6%
—Kimi K2.585.0%
—Gemini 3 Pro83.0%
—Claude Opus 4.581.7%
—LFM3-VL-2B71.3%