Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Abstract Visual Reasoning on ARC-AGI 2
Loading...
100
Accuracy (Pass@2)
best.human
-3.064
23.693
50.45
77.207
Feb 2, 2026
Accuracy (Pass@2)
Updated 4d ago
Evaluation Results
Method
Method
Links
Accuracy (Pass@2)
best.human
2026.02
100
Bespoke (Grok-4)
#Params=1.7T
2026.02
29.4
Grok-4-thinking
#Params=1.7T
2026.02
16
Loop-ViT (Large)
#Params=18M
2026.02
14.2
Loop-ViT (Medium)
#Params=11.2M
2026.02
11.5
VARC (ensemble)
#Params=73M
2026.02
11.1
Loop-ViT (Small)
#Params=3.8M
2026.02
10
VARC
#Params=18M
2026.02
8.3
TRM
#Params=7M
2026.02
7.8
HRM
#Params=27M
2026.02
5
o3-mini-high
2026.02
3
GPT-5
2026.02
1.9
Deepseek R1
#Params=671B
2026.02
1.3
Claude 3.7 8k
2026.02
0.9
Feedback
Search any
task
Search any
task