← Back to community benchmarks

gpt-oss-20b-MXFP4-Q8

M3 Max (30c) · 36 GB · 4bit · 2026-03-11

Performance

1k

tokens

642.1

PP tok/s

72.4

TG tok/s

1595

TTFT (ms)

11.4

Peak mem (GB)

Hardware

Chip M3 Max (30c)

Memory 36 GB

GPU Cores 30

Software

oMLX v0.2.8

macOS macOS 26.2

Context 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	642.1	72.4	11.4 GB	current
4k	741.5	66.9	11.5 GB	view

Batching Results

Batch Size	TG tok/s	Speedup
1×	72.4	1.00×
2×	100.9	1.39×
4×	124.6	1.72×