← Back to community benchmarks
Qwen3.6-27B4bit-MTP
Performance
1k
tokens
436.4
PP tok/s
16.6
TG tok/s
2346
TTFT (ms)
15.8
Peak mem (GB)
Hardware
Chip
M5 Pro (20c)
Memory
64 GB
GPU Cores
20
Software
oMLX
v0.3.8
macOS
macOS 26.5
Context
1,024
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 436.4 | 16.6 | 15.8 GB | current |
| 4k | 474.6 | 11.0 | 17.3 GB | view |
| 8k | 472.0 | 16.1 | 18.3 GB | view |
| 16k | 453.6 | 9.0 | 19.8 GB | view |
| 32k | 410.8 | 10.3 | 22.8 GB | view |
| 64k | 338.9 | 10.0 | 28.8 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 16.6 | 1.00× |
| 2× | 29.1 | 1.75× |
| 4× | 46.6 | 2.81× |