← Back to community benchmarks

Qwen3.5-27B7bit

M1 Ultra (48c) · 64 GB · 6bit · 2026-03-12
Performance
1k
tokens
125.9
PP tok/s
18.2
TG tok/s
8134
TTFT (ms)
23.6
Peak mem (GB)
Hardware
Chip M1 Ultra (48c)
Memory 64 GB
GPU Cores 48
Software
oMLX v0.2.9
macOS macOS 26.3.1
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 125.9 18.2 23.6 GB current
4k 130.0 17.7 23.9 GB view
8k 129.7 16.9 24.4 GB view
Batching Results
Batch Size TG tok/s Speedup
18.2 1.00×
22.5 1.24×
24.9 1.37×