← Back to community benchmarks

gemma-4-26b-a4b-it

M5 Pro (20c) · 24 GB · 4bit · 2026-05-16
Performance
16k
tokens
1,034
PP tok/s
41.1
TG tok/s
15850
TTFT (ms)
15.6
Peak mem (GB)
Hardware
Chip M5 Pro (20c)
Memory 24 GB
GPU Cores 20
Software
oMLX v0.3.9.dev2
macOS macOS 26.5
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 840.5 49.1 14.3 GB view
4k 1,036 47.5 14.9 GB view
8k 1,073 45.3 15.1 GB view
16k 1,034 41.1 15.6 GB current
32k 928.6 34.2 16.5 GB view
64k 636.3 24.9 18.6 GB view