List your company
ResearchFormerly WizWand

Vision Reasoning on Tunnel Vision, BabyVision, and TRACER-Bench

98.3Overall ScoreHuman
21.13241.16661.281.234May 13, 2026
Updated 5mo ago

Evaluation results

MethodLinks
Human2026.0598.3
GPT-5.22026.0561.8
Gemini 3.1 Pro2026.0556.4
CAVE-7BParam=7B, Instruct=false2026.0533.3
CAVE-3BParam=3B, Instruct=false2026.0530.5
Qwen3-VLParam=8B, Instruct=true2026.0530.1
CAVE Belief-OnlyParam=7B, Instruct=false2026.0529.6
InternVL3.5Param=8B, Instruct=true2026.0529.2
MiniCPM-V 4.5Param=8B, Instruct=false2026.0529.2
DeepEyesParam=7B, Instruct=false2026.0528.5
ThinkLite-VLParam=7B, Instruct=false2026.0527.4
VLAA-ThinkerParam=7B, Instruct=false2026.0527
R1-OneVisionParam=7B, Instruct=false2026.0526
Qwen2.5-VLParam=7B, Instruct=true2026.0525.9
Random2026.0524.1