← Back to community benchmarks
Qwen3.6-27B-oQ8-mtp
Performance
16k
tokens
648.9
PP tok/s
24.8
TG tok/s
25250
TTFT (ms)
32.8
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.9
macOS
macOS 26.5
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 728.0 | 30.6 | 28.8 GB | view |
| 4k | 738.9 | 29.2 | 30.3 GB | view |
| 8k | 682.4 | 27.3 | 31.3 GB | view |
| 16k | 648.9 | 24.8 | 32.8 GB | current |
| 32k | 598.3 | 22.8 | 35.8 GB | view |
| 64k | 483.7 | 18.0 | 41.8 GB | view |
| 128k | 291.8 | 13.3 | 54.2 GB | view |
| 195k | 225.1 | 2.3 | 74.6 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 30.6 | 1.00× |
| 2× | 25.9 | 0.85× |
| 4× | 38.0 | 1.24× |