← Back to community benchmarks

gemma-4-e4b-it-mxfp8

M4 (10c) · 24 GB · 8bit · 2026-05-19
Performance
16k
tokens
487.0
PP tok/s
17.2
TG tok/s
33641
TTFT (ms)
9.5
Peak mem (GB)
Hardware
Chip M4 (10c)
Memory 24 GB
GPU Cores 10
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 537.0 19.2 8.7 GB view
4k 524.1 19.1 9.0 GB view
8k 511.0 18.5 9.1 GB view
16k 487.0 17.2 9.5 GB current
32k 367.0 15.1 10.3 GB view
64k 284.8 11.8 12.0 GB view
128k 216.9 9.1 15.4 GB view
Batching Results
Batch Size TG tok/s Speedup
19.2 1.00×
24.1 1.26×
27.4 1.43×
28.0 1.46×