List your company
ResearchFormerly WizWand

Real-world perception-centric reasoning on V* (test)

84.25AccuracyGLM-9B-DeltaThinker
57.074864.129971.18578.2401May 15, 2026
Updated 5mo ago

Evaluation results

MethodLinks
GLM-9B-DeltaThinkerAlgorithm=DeltaPrompts2026.0584.25
Qwen-8B-DeltaThinkerAlgorithm=DeltaPrompts2026.0582.73
Vision-R1-7B2026.0581.15
GLM-4.1V-9B-Thinking2026.0579.58
Qwen3-VL-8B-Thinking2026.0576.96
ARES-RL-7B2026.0571.2
REVisual-R12026.0569.11
Bee-8B-RL2026.0558.12