← Back to community benchmarks

gpt-oss-120b-MXFP4-Q8

M4 Max (40c) · 128 GB · 4bit · 2026-04-04

Performance

16k

tokens

734.3

PP tok/s

48.8

TG tok/s

22314

TTFT (ms)

60.0

Peak mem (GB)

Hardware

Chip M4 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.3.2

macOS macOS 26.4

Context 16,384

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	785.4	73.1	59.5 GB	view
4k	852.0	67.9	59.6 GB	view
8k	825.1	62.2	59.8 GB	view
16k	734.3	48.8	60.0 GB	current
32k	627.7	34.8	60.6 GB	view
64k	521.5	22.1	61.7 GB	view