← Back to community benchmarks

Qwen3.6-27BoQ8

M3 Ultra (80c) · 512 GB · 8bit · 2026-05-04
Performance
4k
tokens
419.6
PP tok/s
21.3
TG tok/s
9763
TTFT (ms)
29.8
Peak mem (GB)
Hardware
Chip M3 Ultra (80c)
Memory 512 GB
GPU Cores 80
Software
oMLX v0.3.8
macOS macOS 26.3
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 387.0 21.5 28.3 GB view
4k 419.6 21.3 29.8 GB current
8k 419.3 20.9 30.8 GB view
16k 408.6 20.4 32.3 GB view
Batching Results
Batch Size TG tok/s Speedup
21.5 1.00×
38.5 1.79×
57.5 2.67×