← Back to community benchmarks

Qwen3.5-122B-A10B

M3 Ultra (60c) · 96 GB · 4bit · 2026-04-16

Performance

1k

tokens

627.3

PP tok/s

56.5

TG tok/s

1632

TTFT (ms)

65.6

Peak mem (GB)

Hardware

Chip M3 Ultra (60c)

Memory 96 GB

GPU Cores 60

Software

oMLX v0.3.4

macOS macOS 26.2

Context 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	627.3	56.5	65.6 GB	current
4k	747.2	52.8	66.9 GB	view
8k	741.8	50.3	67.5 GB	view
16k	700.8	41.4	68.7 GB	view
32k	614.2	32.0	71.1 GB	view
64k	428.8	21.7	76.0 GB	view