← Back to community benchmarks

Qwen3.5-27B-Claude-4.6-Opus-Distilled

M3 Ultra (80c) · 256 GB · 4bit · 2026-06-22

Performance

1k

tokens

396.2

PP tok/s

41.4

TG tok/s

2584

TTFT (ms)

15.8

Peak mem (GB)

Hardware

Chip M3 Ultra (80c)

Memory 256 GB

GPU Cores 80

Software

oMLX v0.4.4rc2

macOS macOS 26.5.1

Context 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	396.2	41.4	15.8 GB	current
4k	424.7	40.5	17.3 GB	view

Batching Results

Batch Size	TG tok/s	Speedup
1×	41.4	1.00×
2×	70.0	1.69×
4×	90.8	2.19×