Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Spectral Inspection on Cross-Task (test)
Loading...
77
Score O
Spec-o3-7B
23.96
37.73
51.5
65.27
Jan 10, 2026
Score O
Score B
Score A
Average Score
Updated 5mo ago
Evaluation Results
Method
Method
Links
Score O
Score B
Score A
Average Score
Spec-o3-7B
Model Size=7B
2026.01
77
73.1
79.1
76.4
Spec-o3-3B
Model Size=3B
2026.01
73.5
72.6
77.1
74.4
o3
2026.01
57.6
64.2
60.8
60.9
GPT-4.1
2026.01
32.2
34.1
33
33.1
Qwen3-VL-30B-A3B
Model Size=30B-A3B
2026.01
32
33.7
32.7
32.8
Qwen3-VL-8B
Model Size=8B
2026.01
30.7
32.2
30.8
31.2
Qwen2.5-VL-7B
Model Size=7B
2026.01
29.4
31.8
30.3
30.5
Qwen2.5-VL-3B
Model Size=3B
2026.01
26
28.1
27.2
27.1
Feedback
Search any
task
Search any
task