List your company

Multi-path Speculative Decoding on Average across models (Qwen, Gemma, Llama) (test)

16.69Throughput (tokens/s)SpecInfer NDE
14.755615.257815.7616.2622Feb 19, 2026

Evaluation results

MethodLinks
SpecInfer NDEExpansion Mode=Neural Dynamic Expansion2026.0216.69
SpecTr NDEExpansion Mode=Neural Dynamic Expansion2026.0216.4
NaiveTree NDEExpansion Mode=Neural Dynamic Expansion2026.0216.18
TraversalExpansion Mode=Static2026.0215.89
NSS NDEExpansion Mode=Neural Dynamic Expansion2026.0215.05
Khisti NDEExpansion Mode=Neural Dynamic Expansion2026.0214.83