← Back to community benchmarks
Qwen3.6-27B
Performance
8k
tokens
437.2
PP tok/s
9.8
TG tok/s
18739
TTFT (ms)
31.1
Peak mem (GB)
Hardware
Chip
M5 Pro (20c)
Memory
64 GB
GPU Cores
20
Software
oMLX
v0.5.1
macOS
macOS 26.5.2
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 424.5 | 10.0 | 28.3 GB | view |
| 4k | 453.0 | 9.9 | 29.8 GB | view |
| 8k | 437.2 | 9.8 | 31.1 GB | current |
| 16k | 398.0 | 9.6 | 32.9 GB | view |
| 32k | 361.4 | 9.2 | 35.6 GB | view |
| 64k | 295.3 | 8.3 | 42.7 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 10.0 | 1.00× |
| 2× | 18.8 | 1.88× |
| 4× | 34.4 | 3.44× |
| 8× | 68.8 | 6.88× |