← Back to community benchmarks
Qwen3.6-27B-oQ6-mtp
Performance
4k
tokens
825.1
PP tok/s
32.4
TG tok/s
4964
TTFT (ms)
24.6
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.9.dev1
macOS
macOS 26.5
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 741.4 | 30.7 | 23.2 GB | view |
| 4k | 825.1 | 32.4 | 24.6 GB | current |
| 8k | 774.0 | 31.6 | 25.7 GB | view |
| 16k | 752.9 | 29.4 | 27.2 GB | view |
| 32k | 695.8 | 28.1 | 30.2 GB | view |
| 64k | 592.3 | 23.8 | 36.2 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 30.7 | 1.00× |
| 2× | 37.8 | 1.23× |
| 4× | 47.8 | 1.56× |