← Back to community benchmarks
Qwen3.6-27B
Performance
1k
tokens
547.6
PP tok/s
28.3
TG tok/s
1870
TTFT (ms)
15.8
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.5.1
macOS
macOS 26.5
Context
1,024
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 547.6 | 28.3 | 15.8 GB | current |
| 4k | 564.5 | 28.0 | 17.3 GB | view |
| 8k | 642.1 | 30.1 | 18.3 GB | view |
| 16k | 618.5 | 27.5 | 20.4 GB | view |
| 32k | 543.4 | 23.8 | 22.8 GB | view |
| 64k | 386.7 | 18.3 | 29.9 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 28.3 | 1.00× |
| 2× | 39.4 | 1.39× |
| 4× | 56.1 | 1.98× |
| 8× | 68.1 | 2.41× |