← Back to community benchmarks

Qwen3.5-122B-A10B-5bit

M1 Ultra (64c) · 128 GB · 5bit · 2026-03-11

Performance

8k

tokens

348.0

PP tok/s

29.8

TG tok/s

23542

TTFT (ms)

81.2

Peak mem (GB)

Hardware

Chip M1 Ultra (64c)

Memory 128 GB

GPU Cores 64

Software

oMLX v0.2.7

macOS macOS 26.4

Context 8,192

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	319.8	32.9	80.6 GB	view
4k	349.1	31.7	80.9 GB	view
8k	348.0	29.8	81.2 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	32.9	1.00×
2×	48.8	1.48×
4×	65.7	2.00×