← Back to community benchmarks

gemma-4-26b-a4b-it

M5 (10c) · 32 GB · 4bit · 2026-05-08
Performance
32k
tokens
832.6
PP tok/s
26.8
TG tok/s
39355
TTFT (ms)
16.4
Peak mem (GB)
Hardware
Chip M5 (10c)
Memory 32 GB
GPU Cores 10
Software
oMLX v0.3.8
macOS macOS 26.4.1
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 781.9 40.4 14.2 GB view
4k 934.9 38.5 14.9 GB view
8k 1,033 36.6 15.0 GB view
16k 983.2 33.1 15.5 GB view
32k 832.6 26.8 16.4 GB current
64k 667.2 19.8 17.7 GB view