← Back to community benchmarks

gemma-4-26B-A4B-it

M3 Ultra (60c) · 96 GB · 4bit · 2026-05-06

Performance

64k

tokens

1,218

PP tok/s

28.3

TG tok/s

53796

TTFT (ms)

17.7

Peak mem (GB)

Hardware

Chip M3 Ultra (60c)

Memory 96 GB

GPU Cores 60

Software

oMLX v0.3.8rc1

macOS macOS 15.7.5

Context 65,536

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,300	99.8	14.3 GB	view
4k	1,567	92.2	14.7 GB	view
8k	1,710	79.1	14.8 GB	view
16k	1,665	63.5	15.2 GB	view
32k	1,506	45.1	16.0 GB	view
64k	1,218	28.3	17.7 GB	current
128k	858.8	8.7	21.2 GB	view