← Back to community benchmarks

gemma-4-31b-it

M3 Ultra (60c) · 256 GB · 8bit · 2026-05-04
Performance
16k
tokens
243.1
PP tok/s
16.8
TG tok/s
67409
TTFT (ms)
36.1
Peak mem (GB)
Hardware
Chip M3 Ultra (60c)
Memory 256 GB
GPU Cores 60
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 248.9 18.8 31.8 GB view
4k 253.3 18.3 34.5 GB view
8k 251.1 17.8 35.0 GB view
16k 243.1 16.8 36.1 GB current
32k 226.9 15.2 39.2 GB view
64k 197.9 12.8 45.1 GB view
128k 153.7 7.8 63.4 GB view
Batching Results
Batch Size TG tok/s Speedup
18.8 1.00×
31.2 1.66×
39.0 2.07×