List your company

Decode Throughput on Gemma 4 E2B

127.7Throughput (tok/s)BaseRT
56.77275.18693.6112.014Jul 1, 2026

Evaluation results

MethodLinks
BaseRTQuantization=Q4, Hardware=Apple M4 Pro, Target Length=tg1282026.07127.71.19—
llama.cppQuantization=Q4, Hardware=Apple M4 Pro, Target Length=tg1282026.07107——
BaseRTQuantization=Q8, Hardware=Apple M4 Pro, Target Length=tg1282026.0784.51.42—
llama.cppQuantization=Q8, Hardware=Apple M4 Pro, Target Length=tg1282026.0759.5——