← Back to community benchmarks

gemma-4-26b-a4b-it-oQ8

M4 Pro (20c) · 64 GB · 8bit · 2026-05-06
Performance
16k
tokens
634.1
PP tok/s
39.5
TG tok/s
25838
TTFT (ms)
27.1
Peak mem (GB)
Hardware
Chip M4 Pro (20c)
Memory 64 GB
GPU Cores 20
Software
oMLX v0.3.8
macOS macOS 26.3
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
4k 307.0 43.1 26.2 GB view
8k 717.3 41.5 26.6 GB view
16k 634.1 39.5 27.1 GB current
32k 560.6 34.5 27.5 GB view
64k 475.9 1.9 36.8 GB view
128k 353.6 1.7 38.0 GB view