← Back to community benchmarks

Qwen3.5-40B-Claude-4.6-Opus-Deckard-Heretic-Uncensored-Thinking

M1 Pro (14c) · 32 GB · 4bit · 2026-03-18
Performance
1k
tokens
27.6
PP tok/s
6.0
TG tok/s
37137
TTFT (ms)
22.5
Peak mem (GB)
Hardware
Chip M1 Pro (14c)
Memory 32 GB
GPU Cores 14
Software
oMLX v0.2.19.dev1
macOS macOS 26.3.1
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 27.6 6.0 22.5 GB current
4k 27.5 6.0 24.6 GB view
Batching Results
Batch Size TG tok/s Speedup
6.0 1.00×
6.5 1.08×
6.4 1.07×