Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
LLM Throughput on AIME (Mathematical Reasoning)
Loading...
998
Throughput (token/s)
FlashEvolve
67.2
308.85
550.5
792.15
May 8, 2026
Throughput (token/s)
Proposal Throughput (proposal/min)
Updated 22d ago
Evaluation Results
Method
Method
Links
Throughput (token/s)
Proposal Throughput (proposal/min)
FlashEvolve
Deployment=vLLM, Model...
2026.05
998
11.4
Combee
Deployment=vLLM, Model...
2026.05
994
6.2
Combee
Deployment=vLLM, Model...
2026.05
977
1.6
FlashEvolve
Deployment=API, Model=...
2026.05
485
6.6
Combee
Deployment=API, Model=...
2026.05
336
0.6
Combee
Deployment=API, Model=...
2026.05
211
1
GEPA
Deployment=vLLM, Model...
2026.05
200
2.2
GEPA
Deployment=API, Model=...
2026.05
103
1.3
Feedback
Search any
task
Search any
task