← Back to community benchmarks

Qwen3.6-35B-A3B-oQ4

M1 Max (24c) · 32 GB · 4bit · 2026-04-17
Performance
16k
tokens
278.2
PP tok/s
30.5
TG tok/s
58898
TTFT (ms)
21.7
Peak mem (GB)
Hardware
Chip M1 Max (24c)
Memory 32 GB
GPU Cores 24
Software
oMLX v0.3.6
macOS macOS 26.4.1
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 270.5 44.6 19.9 GB view
4k 309.8 43.1 20.7 GB view
8k 282.6 35.3 21.1 GB view
16k 278.2 30.5 21.7 GB current
Batching Results
Batch Size TG tok/s Speedup
44.6 1.00×
50.4 1.13×
56.4 1.26×