← Back to community benchmarks

Qwen3.6-35B-A3B-TurboQuant

M5 (10c) · 32 GB · 4bit · 2026-05-01
Performance
1k
tokens
549.7
PP tok/s
47.5
TG tok/s
1863
TTFT (ms)
19.2
Peak mem (GB)
Hardware
Chip M5 (10c)
Memory 32 GB
GPU Cores 10
Software
oMLX v0.3.8
macOS macOS 26.3.1
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 549.7 47.5 19.2 GB current
4k 684.0 43.9 19.9 GB view
8k 696.2 42.2 20.4 GB view
16k 646.8 37.0 21.1 GB view
32k 538.8 30.3 22.6 GB view
Batching Results
Batch Size TG tok/s Speedup
47.5 1.00×
58.6 1.23×
68.8 1.45×