Multimodal Reasoning on MMVet v1 (val) benchmark leaderboard · SOTA2 Research

Multimodal Reasoning on MMVet v1 (val)

33.7AccuracyEvoPrune

Evaluation Results

Method	Links
EvoPrune2026.03		33.7	1.1	0.77	0.84	74.9	97.9
FasterVLM2026.03		33.1	1.15	0.77	0.85	72.3	94.6
VisPruner2026.03		32.8	1.32	0.77	0.9	73.2	95.7
EvoPrune2026.03		32.4	1.07	0.73	0.77	72.9	95.3
Native2026.03		32.2	1	1	1	76.5	100
CDPruner2026.03		31.7	1.41	0.77	0.91	73.9	96.6
DivPrune2026.03		31.5	1.16	0.77	0.85	72.9	95.3
DART2026.03		31.3	1.2	0.77	0.88	70.9	92.7
CDPruner2026.03		30.3	1.38	0.73	0.86	72.2	94.4
DivPrune2026.03		29.9	1.12	0.73	0.79	70.4	92
VisPruner2026.03		29.5	1.35	0.73	0.86	69.9	91.3
DART2026.03		28.8	1.13	0.73	0.8	66.2	86.6
CDPruner2026.03		28.6	1.22	0.71	0.8	69.6	91
DivPrune2026.03		26.9	1.1	0.71	0.77	66.8	87.4
FasterVLM2026.03		26.8	1.09	0.73	0.79	68.2	89.2
VisPruner2026.03		26.6	1.35	0.71	0.84	65.2	85.2
EvoPrune2026.03		26.6	1.06	0.71	0.75	68.7	89.8
FasterVLM2026.03		26.5	1.09	0.71	0.77	64.5	84.4
DART2026.03		23.7	1.09	0.71	0.76	59.8	78.1