← Back to community benchmarks

gemma-4-e4b-it

M3 Ultra (60c) · 256 GB · 4bit · 2026-05-21
Performance
16k
tokens
2,754
PP tok/s
81.1
TG tok/s
5948
TTFT (ms)
5.9
Peak mem (GB)
Hardware
Chip M3 Ultra (60c)
Memory 256 GB
GPU Cores 60
Software
oMLX v0.3.9rc1
macOS macOS 26.4.1
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 2,849 100.8 5.6 GB view
4k 2,948 101.2 5.7 GB view
8k 2,872 94.7 5.8 GB view
16k 2,754 81.1 5.9 GB current
32k 2,502 65.3 6.4 GB view
64k 2,137 45.9 7.1 GB view
Batching Results
Batch Size TG tok/s Speedup
100.8 1.00×
187.6 1.86×
251.0 2.49×
338.5 3.36×