← Back to community benchmarks

Qwen3.6-35B-A3B-Ornith-80-20-Beta-oQ8e-mtp

M5 Max (40c) · 128 GB · 8bit · 2026-07-18
Performance
1k
tokens
3,463
PP tok/s
99.5
TG tok/s
296
TTFT (ms)
35.5
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.5.2.dev1
macOS macOS 27.0
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 3,463 99.5 35.5 GB current
4k 4,177 95.6 36.3 GB view
32k 3,355 84.0 38.7 GB view
64k 2,426 74.9 41.9 GB view
128k 1,536 60.5 47.9 GB view
Batching Results
Batch Size TG tok/s Speedup
99.5 1.00×
157.1 1.58×
241.8 2.43×