← Back to community benchmarks

Llama-3.2-1B-Instruct

M4 (10c) · 32 GB · 4bit · 2026-04-26

Performance

8k

tokens

1,127

PP tok/s

60.7

TG tok/s

7268

TTFT (ms)

1.4

Peak mem (GB)

Hardware

Chip M4 (10c)

Memory 32 GB

GPU Cores 10

Software

oMLX v0.3.7rc1

macOS macOS 26.3.1

Context 8,192

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,274	113.2	1.3 GB	view
4k	1,249	81.7	1.4 GB	view
8k	1,127	60.7	1.4 GB	current
16k	912.0	58.3	1.7 GB	view