← Back to community benchmarks

Gemma-4-26B-A4B4.7bit-vision

M5 Pro (20c) · 64 GB · 4bit · 2026-07-05

Performance

4k

tokens

1,853

PP tok/s

71.3

TG tok/s

2210

TTFT (ms)

15.1

Peak mem (GB)

Hardware

Chip M5 Pro (20c)

Memory 64 GB

GPU Cores 20

Software

oMLX v0.4.4

macOS macOS 26.5.2

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,551	74.2	14.6 GB	view
4k	1,853	71.3	15.1 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	74.2	1.00×
2×	102.4	1.38×
4×	270.8	3.65×