← Back to community benchmarks

gemma-4-31b-it

M5 Max (40c) · 128 GB · 4bit · 2026-06-01
Performance
8k
tokens
452.1
PP tok/s
17.5
TG tok/s
18122
TTFT (ms)
20.7
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.12
macOS macOS 26.4.1
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 661.9 24.7 17.6 GB view
4k 558.8 19.3 20.2 GB view
8k 452.1 17.5 20.7 GB current
16k 406.4 14.8 21.9 GB view
32k 304.6 9.5 24.9 GB view
64k 230.8 8.0 30.9 GB view
Batching Results
Batch Size TG tok/s Speedup
24.7 1.00×
19.5 0.79×
24.6 1.00×