← Back to community benchmarks

Qwen3.5-27B-heretic

M3 Max (40c) · 64 GB · 8bit · 2026-03-11
Performance
4k
tokens
194.7
PP tok/s
12.5
TG tok/s
21032
TTFT (ms)
29.5
Peak mem (GB)
Hardware
Chip M3 Max (40c)
Memory 64 GB
GPU Cores 40
Software
oMLX v0.2.7
macOS macOS 26.3.1
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 211.1 12.9 29.1 GB view
4k 194.7 12.5 29.5 GB current
Batching Results
Batch Size TG tok/s Speedup
12.9 1.00×
21.0 1.63×
30.2 2.34×