← Back to community benchmarks

gemma-4-e4b-it

M4 (10c) · 24 GB · 8bit · 2026-05-04
Performance
195k
tokens
221.5
PP tok/s
8.0
TG tok/s
902766
TTFT (ms)
13.2
Peak mem (GB)
Hardware
Chip M4 (10c)
Memory 24 GB
GPU Cores 10
Software
oMLX v0.3.6
macOS macOS 26.4.1
Context 200,000
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 417.0 19.5 7.9 GB view
4k 460.6 19.0 8.0 GB view
8k 474.5 18.3 8.1 GB view
16k 454.3 16.7 8.3 GB view
32k 412.0 14.5 8.7 GB view
64k 257.9 8.5 9.6 GB view
128k 216.2 9.8 11.3 GB view
195k 221.5 8.0 13.2 GB current
Batching Results
Batch Size TG tok/s Speedup
19.5 1.00×
36.5 1.87×
48.3 2.48×
52.0 2.67×