List your company

Token Throughput on LLaMA 8B 3.1 (4K-128K Context)

150.8Token Throughput (4K Context) (tokens/sec)ShadowKV
134.888139.019143.15147.281Mar 27, 2026

Evaluation results

MethodLinks
ShadowKVBatch size=82026.03150.8136.7105.267.534.9
TTKVBatch size=82026.03148.7142.112895.365.7
KVQuant (3-bit)Batch size=82026.03145.1130.598.66130.8
KIVI (2-bit)Batch size=82026.03142.31289558.729.4
FP16Batch size=82026.03135.5120.489.2——