← Back to community benchmarks

gemma-4-12B-it-qat

M4 Max (40c) · 64 GB · 4bit · 2026-06-06

Performance

8k

tokens

504.7

PP tok/s

34.4

TG tok/s

16232

TTFT (ms)

13.0

Peak mem (GB)

Hardware

Chip M4 Max (40c)

Memory 64 GB

GPU Cores 40

Software

oMLX v0.4.2.dev3

macOS macOS 26.5

Context 8,192

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
4k	513.2	34.8	12.5 GB	view
8k	504.7	34.4	13.0 GB	current
16k	482.4	33.6	13.1 GB	view