Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Language Understanding on MMLU-Pro (TPS and Latency)
Loading...
25.8
TPS
MoE-SpAc
0.1432
6.8041
13.465
20.1259
Feb 12, 2026
TPS
Latency
Updated 1mo ago
Evaluation Results
Method
Method
Links
TPS
Latency
MoE-SpAc
2026.02
25.8
23.11
llama.cpp
speculative decoding=true
2026.02
17.29
32.88
llama.cpp
speculative decoding=f...
2026.02
15.17
37.92
HybriMoE
2026.02
12.22
47.62
Fate
2026.02
7.93
73.65
SP-MoE
2026.02
4.75
116.73
MoE-Infinity
2026.02
3.64
151.43
vLLM
2026.02
2.65
203.74
Mixtral Offload
2026.02
2.43
230.87
Accelerate
2026.02
1.13
495.37
Feedback
Search any
task
Search any
task