← Back to community benchmarks

DeepSeek-R1-0528-Qwen3-8B

M4 Pro (16c) · 48 GB · 4bit · 2026-03-10

Performance

1k

tokens

350.0

PP tok/s

49.4

TG tok/s

2926

TTFT (ms)

5.0

Peak mem (GB)

Hardware

Chip M4 Pro (16c)

Memory 48 GB

GPU Cores 16

Software

oMLX v0.2.7

macOS macOS 26.2

Context 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	350.0	49.4	5.0 GB	current
4k	335.6	42.5	5.6 GB	view

Batching Results

Batch Size	TG tok/s	Speedup
1×	49.4	1.00×
2×	76.4	1.55×
4×	81.8	1.66×