← Back to community benchmarks

gemma-4-31b-it

M3 Max (40c) · 128 GB · bf16 · 2026-05-26
Performance
32k
tokens
131.6
PP tok/s
3.8
TG tok/s
248939
TTFT (ms)
63.8
Peak mem (GB)
Hardware
Chip M3 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.10
macOS macOS 26.3.1
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 184.5 5.9 58.5 GB view
4k 187.9 5.8 59.8 GB view
8k 176.0 5.8 60.3 GB view
16k 149.8 3.4 61.5 GB view
32k 131.6 3.8 63.8 GB current
64k 112.3 3.5 68.5 GB view
128k 81.6 2.9 78.1 GB view
Batching Results
Batch Size TG tok/s Speedup
5.9 1.00×
4.8 0.81×
9.5 1.61×