← Back to community benchmarks

gemma-4-E4B-it

M3 Pro (18c) · 36 GB · 4bit · 2026-05-09
Performance
16k
tokens
932.9
PP tok/s
25.5
TG tok/s
17563
TTFT (ms)
7.5
Peak mem (GB)
Hardware
Chip M3 Pro (18c)
Memory 36 GB
GPU Cores 18
Software
oMLX v0.3.8
macOS macOS 26.4.1
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 962.9 28.0 7.0 GB view
4k 985.3 27.3 7.1 GB view
8k 970.2 26.9 7.3 GB view
16k 932.9 25.5 7.5 GB current
32k 862.3 23.1 7.9 GB view
64k 748.0 19.6 8.6 GB view
Batching Results
Batch Size TG tok/s Speedup
28.0 1.00×
56.2 2.01×
97.9 3.50×