← Back to community benchmarks

Qwen3.5-4B-8bit

M4 (10c) · 16 GB · 8bit · 2026-03-10

Performance

1k

tokens

402.9

PP tok/s

22.3

TG tok/s

2541

TTFT (ms)

5.8

Peak mem (GB)

Hardware

Chip M4 (10c)

Memory 16 GB

GPU Cores 10

Software

oMLX v0.2.6

macOS macOS 26.2

Context 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	402.9	22.3	5.8 GB	current
4k	403.2	21.3	6.0 GB	view

Batching Results

Batch Size	TG tok/s	Speedup
1×	22.3	1.00×
2×	43.5	1.95×
4×	69.6	3.12×