← Back to community benchmarks

gpt-oss-20b-MXFP4-Q8

M4 (10c) · 32 GB · 4bit · 2026-03-19

Performance

4k

tokens

324.0

PP tok/s

29.5

TG tok/s

12642

TTFT (ms)

11.5

Peak mem (GB)

Hardware

Chip M4 (10c)

Memory 32 GB

GPU Cores 10

Software

oMLX v0.2.13

macOS macOS 26.3.1

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	310.9	33.8	11.4 GB	view
4k	324.0	29.5	11.5 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	33.8	1.00×
2×	44.0	1.30×
4×	50.6	1.50×