← Back to community benchmarks

Huihui-Qwen3.5-35B-A3B-abliterated-8bit

M3 Ultra (80c) · 512 GB · 8bit · 2026-03-15

Performance

1k

tokens

2,514

PP tok/s

73.6

TG tok/s

407

TTFT (ms)

36.2

Peak mem (GB)

Hardware

Chip M3 Ultra (80c)

Memory 512 GB

GPU Cores 80

Software

oMLX v0.2.10

macOS macOS 15.3

Context 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	2,514	73.6	36.2 GB	current
4k	2,642	71.6	36.3 GB	view

Batching Results

Batch Size	TG tok/s	Speedup
1×	73.6	1.00×
2×	129.9	1.76×
4×	216.1	2.94×