← Back to community benchmarks
Qwen3.6-27B-oQ8-mtp
Performance
32k
tokens
701.8
PP tok/s
25.8
TG tok/s
46691
TTFT (ms)
36.2
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.9.dev2
macOS
macOS 26.5
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 723.7 | 29.6 | 29.3 GB | view |
| 4k | 838.7 | 28.0 | 30.7 GB | view |
| 8k | 790.7 | 28.7 | 31.7 GB | view |
| 16k | 767.0 | 27.1 | 33.2 GB | view |
| 32k | 701.8 | 25.8 | 36.2 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 29.6 | 1.00× |
| 2× | 34.9 | 1.18× |
| 4× | 50.6 | 1.71× |