MMBench Benchmark Leaderboard | Phaseo

MMBench

AI benchmark results and model performance

Type: percentageMultimodal

Recorded Results

14

Average Score

83.49%

Score Range

69.20% - 88%

Leading Model (lowest score)

69.20% - DeepSeek VL2 Tiny

Scores Over Time

Individual benchmark scores plotted by date.

Models Using This Benchmark

Organisation	Model	Reported	Top Score	Info	Self Reported	Source
DeepSeek	DeepSeek VL2 Tiny	13 Dec 2024	69.20%	-	Yes	Source
DeepSeek	DeepSeek VL2	13 Dec 2024	79.60%	-	Yes	Source
DeepSeek	DeepSeek VL2 Small	13 Dec 2024	80.30%	-	Yes	Source
Microsoft	Phi 3 Vision 128K Instruct	-	81.90%	inferred family alias from phi-3.5-vision-instruct (score=0.3495; benches=9)	Yes	Source
Microsoft	Phi 3.5 vision instruct	23 Aug 2024	81.90%	-	Yes	Source
Qwen	Qwen 2.5 0.5B	-	84.30%	inferred family alias from qwen2.5-vl-7b (score=0.3131; benches=32)	Yes	Source
Qwen	Qwen 2.5 1.5B	-	84.30%	inferred family alias from qwen2.5-vl-7b (score=0.3131; benches=32)	Yes	Source
Qwen	Qwen 2.5 3B	-	84.30%	inferred family alias from qwen2.5-vl-7b (score=0.3531; benches=32)	Yes	Source
Qwen	Qwen 2.5 VL 7B Instruct	-	84.30%	-	Yes	Source
Microsoft	Phi 4 multimodal instruct	01 Feb 2025	86.70%	-	Yes	Source
Qwen	Qwen 2.5 Math 72B	-	88%	inferred high-confidence family alias from qwen2.5-vl-72b (score=0.4700; benches=30)	Yes	Source
Qwen	Qwen 2.5 Math PRM 72B	-	88%	inferred family alias from qwen2.5-vl-72b (score=0.4050; benches=30)	Yes	Source
Qwen	Qwen 2.5 Math RM 72B	-	88%	inferred family alias from qwen2.5-vl-72b (score=0.4092; benches=30)	Yes	Source
Qwen	Qwen 2.5 VL 72B Instruct	26 Jan 2025	88%	-	Yes	Source