← Back to community benchmarks

Qwen3.5-27B-Claude-4.6-Opus-Distilled

M4 Max (40c) · 64 GB · 4bit · 2026-04-25
Performance
32k
tokens
145.3
PP tok/s
14.9
TG tok/s
225521
TTFT (ms)
21.6
Peak mem (GB)
Hardware
Chip M4 Max (40c)
Memory 64 GB
GPU Cores 40
Software
oMLX v0.3.8.dev1
macOS macOS 15.7.4
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 240.2 29.0 15.8 GB view
4k 214.6 26.1 17.3 GB view
8k 197.6 25.6 17.9 GB view
16k 185.6 18.0 19.1 GB view
32k 145.3 14.9 21.6 GB current
Batching Results
Batch Size TG tok/s Speedup
29.0 1.00×
30.8 1.06×
36.6 1.26×
41.3 1.42×