← Back to community benchmarks

Llama-3.2-3B-Instruct

M4 (10c) · 16 GB · 4bit · 2026-03-12

Performance

1k

tokens

551.4

PP tok/s

47.7

TG tok/s

1859

TTFT (ms)

2.3

Peak mem (GB)

Hardware

Chip M4 (10c)

Memory 16 GB

GPU Cores 10

Software

oMLX v0.2.7

macOS macOS 15.6

Context 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	551.4	47.7	2.3 GB	current
4k	517.6	38.9	2.6 GB	view

Batching Results

Batch Size	TG tok/s	Speedup
1×	47.7	1.00×
2×	88.8	1.86×
4×	115.0	2.41×