Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Question Answering on GPQA Diamond (Quality, Cost, and Error Reduction)
Loading...
100
Quality (%)
Okshot
40.096
55.648
71.2
86.752
Jun 25, 2026
Jun 26, 2026
Jun 27, 2026
Jun 28, 2026
Quality (%)
Cost ($)
Error Reduction (%)
Updated 25d ago
Evaluation Results
Method
Method
Links
Quality (%)
Cost ($)
Error Reduction (%)
Okshot
k=1
2026.06
100
0.036
100
SOTA LLM
2026.06
94
0.005
-
Qwen3-235B
Category=Heuristic
2026.06
78.8
1.452
-
ENTROROUTER
2026.06
73.2
0.902
-
FORC
Category=Existing Rout...
2026.06
71.7
0.741
-
Random
Category=Heuristic
2026.06
71.2
0.696
-
Qwen3-30B
Category=Heuristic
2026.06
69.7
0.707
-
RouteLLM
Category=Existing Rout...
2026.06
65.2
0.271
-
Qwen3-4B
Category=Heuristic
2026.06
62.6
0.052
-
Router-R1
Category=Existing Rout...
2026.06
49.5
0.001
-
xRouter
Category=Existing Rout...
2026.06
42.4
0.005
-
Feedback
Search any
task
Search any
task