← Back to community benchmarks
Qwen3.6-27B-oQ6-mtp
Performance
64k
tokens
329.4
PP tok/s
14.6
TG tok/s
198968
TTFT (ms)
36.0
Peak mem (GB)
Hardware
Chip
M5 Pro (20c)
Memory
48 GB
GPU Cores
20
Software
oMLX
v0.3.9
macOS
macOS 26.4.1
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 402.2 | 19.9 | 23.0 GB | view |
| 4k | 444.5 | 19.8 | 24.4 GB | view |
| 8k | 444.3 | 19.0 | 25.4 GB | view |
| 16k | 425.3 | 17.9 | 26.9 GB | view |
| 32k | 393.5 | 16.8 | 29.9 GB | view |
| 64k | 329.4 | 14.6 | 36.0 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 19.9 | 1.00× |
| 2× | 39.3 | 1.97× |
| 4× | 77.1 | 3.87× |