← Back to community benchmarks

Youssofal_Qwen3.6-35B-A3B-Abliterated-Heretic

M5 Max (40c) · 128 GB · 4bit · 2026-07-25
Performance
1k
tokens
3,696
PP tok/s
11.5
TG tok/s
277
TTFT (ms)
49.1
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.5.3
macOS macOS 26.5.1
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 3,696 11.5 49.1 GB current
4k 4,198 73.4 48.1 GB view
8k 4,089 72.4 48.4 GB view
16k 3,561 71.6 49.1 GB view
32k 2,996 67.0 50.4 GB view
64k 2,158 83.7 52.3 GB view
Batching Results
Batch Size TG tok/s Speedup
11.5 1.00×
203.9 17.73×
412.9 35.90×
284.7 24.76×