List your company
ResearchFormerly WizWand

Multimodal Understanding on MMMU-Pro 1.0 (test)

44.39Overall ScoreVanilla (Full Tokens)
17.6124.562531.51538.4675Apr 14, 2026
Updated 6mo ago

Evaluation results

MethodLinks
Vanilla (Full Tokens)Model=InternVL3.5-14B, Model Category=Large-scale MLLMs, Token Retention Ratio=100%2026.0444.39
Vanilla (Full Tokens)Model=Qwen3-VL-8B, Model Category=Large-scale MLLMs, Token Retention Ratio=100%2026.0442.31
DSTPModel=InternVL3.5-14B, Model Category=Large-scale MLLMs, Token Retention Ratio=33.3%2026.0440.64
Vanilla (Full Tokens)Model=Qwen3-VL-4B-Thinking, Model Category=Thinking-based MLLMs, Token Retention Ratio=100%2026.0440.51
Vanilla (Full Tokens)Model=InternVL3.5-8B-Thinking, Model Category=Thinking-based MLLMs, Token Retention Ratio=100%2026.0437.68
DSTPModel=Qwen3-VL-4B-Thinking, Model Category=Thinking-based MLLMs, Token Retention Ratio=33.3%2026.0436.52
DSTPModel=Qwen3-VL-8B, Model Category=Large-scale MLLMs, Token Retention Ratio=33.3%2026.0433.31
FastV (33.3%)Model=InternVL3.5-14B, Model Category=Large-scale MLLMs, Token Retention Ratio=33.3%2026.0431.66
DSTPModel=InternVL3.5-8B-Thinking, Model Category=Thinking-based MLLMs, Token Retention Ratio=33.3%2026.0431.4
FastV (33.3%)Model=Qwen3-VL-4B-Thinking, Model Category=Thinking-based MLLMs, Token Retention Ratio=33.3%2026.0429.77
FastV (33.3%)Model=InternVL3.5-8B-Thinking, Model Category=Thinking-based MLLMs, Token Retention Ratio=33.3%2026.0420.63
FastV (33.3%)Model=Qwen3-VL-8B, Model Category=Large-scale MLLMs, Token Retention Ratio=33.3%2026.0418.64