Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Throughput and Latency Evaluation on MT-Bench
Loading...
24.1
Throughput (TPS)
MoE-SpAc
0.232
6.4285
12.625
18.8215
Feb 12, 2026
Throughput (TPS)
Latency
Updated 1mo ago
Evaluation Results
Method
Method
Links
Throughput (TPS)
Latency
MoE-SpAc
2026.02
24.1
22.26
llama.cpp
speculative decoding=true
2026.02
17.58
29.03
llama.cpp
speculative decoding=f...
2026.02
14.34
36.08
HybriMoE
2026.02
12.03
43.47
Fate
2026.02
7.89
66.89
SP-MoE
2026.02
4.46
111.34
MoE-Infinity
2026.02
3.64
136.84
vLLM
2026.02
2.67
187.88
Mixtral Offload
2026.02
2.36
212.7
Accelerate
2026.02
1.15
435.55
Feedback
Search any
task
Search any
task