← Back to community benchmarks
Qwen3.6-27B-PARO
Performance
8k
tokens
821.6
PP tok/s
33.0
TG tok/s
9971
TTFT (ms)
17.4
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
64 GB
GPU Cores
40
Software
oMLX
v0.4.4
macOS
macOS 26.5.1
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 554.8 | 34.4 | 15.5 GB | view |
| 4k | 812.2 | 33.7 | 16.7 GB | view |
| 8k | 821.6 | 33.0 | 17.4 GB | current |
| 16k | 697.2 | 31.6 | 18.6 GB | view |
| 32k | 643.5 | 29.9 | 21.1 GB | view |
| 64k | 548.8 | 25.6 | 26.2 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 34.4 | 1.00× |
| 2× | 44.8 | 1.30× |
| 4× | 131.6 | 3.83× |