← Back to community benchmarks

Qwen3.6-35B-A3B

M5 Max (40c) · 128 GB · bf16 · 2026-04-16
Performance
16k
tokens
3,362
PP tok/s
61.8
TG tok/s
4873
TTFT (ms)
67.4
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
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 1,466 67.8 65.6 GB view
4k 2,729 67.4 66.4 GB view
8k 3,395 65.1 66.7 GB view
16k 3,362 61.8 67.4 GB current
32k 2,832 52.6 68.7 GB view
64k 1,970 42.3 71.3 GB view
128k 1,290 24.4 76.7 GB view
195k 946.7 17.2 82.5 GB view
Batching Results
Batch Size TG tok/s Speedup
67.8 1.00×
74.8 1.10×
120.1 1.77×
125.3 1.85×