← Back to community benchmarks

gemma-4-26B-A4B-it-OptiQ

M5 (10c) · 32 GB · 4bit · 2026-04-28
Performance
16k
tokens
817.6
PP tok/s
25.3
TG tok/s
20038
TTFT (ms)
16.2
Peak mem (GB)
Hardware
Chip M5 (10c)
Memory 32 GB
GPU Cores 10
Software
oMLX v0.3.6
macOS macOS 26.4.1
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 743.4 37.5 15.3 GB view
4k 847.9 35.3 15.7 GB view
8k 889.7 30.6 15.7 GB view
16k 817.6 25.3 16.2 GB current
32k 608.3 12.9 17.0 GB view
64k 376.8 4.5 18.7 GB view