Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Abstract Reasoning on ARC-AGI v1 (test)
Loading...
98
Accuracy
Best Human
12.512
34.706
56.9
79.094
May 19, 2026
May 21, 2026
May 24, 2026
May 26, 2026
May 29, 2026
May 31, 2026
Jun 3, 2026
Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
Best Human
2026.05
98
Gemini 3 Pro
#Params=N/A
2026.05
75
Grok-4-thinking
#Params=1.7T
2026.05
66.7
Avg. Human
2026.05
60.2
GPT 5.2 (low)
#Params=N/A
2026.05
55.7
GRAM
#Params=10M, Supervisi...
2026.05
52
TRM
#Params=7M, Supervisio...
2026.05
44.6
SePO-Generalist
Task Agent=DeepSeek-V3...
2026.06
43.39
HRM
#Params=27M, Supervisi...
2026.05
40.3
SePO-Specialist
Task Agent=DeepSeek-V3...
2026.06
37.46
Manual-CoT
Task Agent=DeepSeek-V3...
2026.06
37.3
MetaSPO
Task Agent=DeepSeek-V3...
2026.06
37.27
TextGrad
Task Agent=DeepSeek-V3...
2026.06
34.75
o3-mini-high
#Params=N/A
2026.05
34.5
Claude 3.7 16k
#Params=N/A
2026.05
28.6
Direct Pred
#Params=27M, Supervisi...
2026.05
21
Deepseek-R1
#Params=671B
2026.05
15.8
Feedback
Search any
task
Search any
task