← Back to community benchmarks

gemma-4-26b-a4b-it

M2 Ultra (76c) · 192 GB · 8bit · 2026-05-20

Performance

32k

tokens

1,097

PP tok/s

52.2

TG tok/s

29884

TTFT (ms)

27.5

Peak mem (GB)

Hardware

Chip M2 Ultra (76c)

Memory 192 GB

GPU Cores 76

Software

oMLX v0.3.9rc1

macOS macOS 26.4.1

Context 32,768

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	960.6	77.4	25.8 GB	view
4k	1,168	73.9	26.1 GB	view
8k	1,192	70.0	26.3 GB	view
16k	1,169	63.0	26.7 GB	view
32k	1,097	52.2	27.5 GB	current
64k	928.1	38.6	29.2 GB	view
128k	640.3	25.1	32.7 GB	view