← Back to community benchmarks

Llama-3.1-8B-Instruct

M4 (10c) · 16 GB · 4bit · 2026-03-16

Performance

4k

tokens

226.0

PP tok/s

19.9

TG tok/s

18127

TTFT (ms)

5.3

Peak mem (GB)

Hardware

Chip M4 (10c)

Memory 16 GB

GPU Cores 10

Software

oMLX v0.2.12

macOS macOS 26.3.1

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	237.8	22.0	4.9 GB	view
4k	226.0	19.9	5.3 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	22.0	1.00×
2×	43.5	1.98×
4×	54.9	2.50×