← Back to community benchmarks

Qwen3-Coder-Next-abliterated-mxfp4-gs32

M4 Max (40c) · 64 GB · 4bit · 2026-03-11

Performance

16k

tokens

1,182

PP tok/s

69.5

TG tok/s

13862

TTFT (ms)

41.5

Peak mem (GB)

Hardware

Chip M4 Max (40c)

Memory 64 GB

GPU Cores 40

Software

oMLX v0.2.7

macOS macOS 26.4

Context 16,384

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
8k	1,250	81.5	41.1 GB	view
16k	1,182	69.5	41.5 GB	current
32k	1,021	51.1	42.4 GB	view
64k	820.2	32.7	44.2 GB	view