← Back to community benchmarks

Qwen3.5-9B-Claude-4.6-HighIQ-INSTRUCT-HERETIC-UNCENSOREDmxfp8

M5 (8c) · 16 GB · 8bit · 2026-04-17

Performance

1k

tokens

523.5

PP tok/s

15.9

TG tok/s

1956

TTFT (ms)

9.5

Peak mem (GB)

Hardware

Chip M5 (8c)

Memory 16 GB

GPU Cores 8

Software

oMLX v0.3.5

macOS macOS 26.3.2

Context 1,024

Batching Results

Batch Size	TG tok/s	Speedup
1×	15.9	1.00×
2×	29.2	1.84×