← Back to community benchmarks

Qwen3.6-35B-A3B-Claude-4.7-Opus-Reasoning-DistilledoQ4-MTP

M5 (10c) · 32 GB · 4bit · 2026-05-18
Performance
16k
tokens
956.3
PP tok/s
54.2
TG tok/s
17133
TTFT (ms)
21.6
Peak mem (GB)
Hardware
Chip M5 (10c)
Memory 32 GB
GPU Cores 10
Software
oMLX v0.3.9.dev2
macOS macOS 26.5
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 822.1 67.1 19.6 GB view
4k 1,058 61.8 20.4 GB view
8k 1,045 60.2 20.8 GB view
16k 956.3 54.2 21.6 GB current
Batching Results
Batch Size TG tok/s Speedup
67.1 1.00×
89.7 1.34×