← Back to community benchmarks
Qwen3.6-27B-oQ4-mtp
Performance
32k
tokens
131.1
PP tok/s
6.6
TG tok/s
249926
TTFT (ms)
23.6
Peak mem (GB)
Hardware
Chip
M5 (10c)
Memory
32 GB
GPU Cores
10
Software
oMLX
v0.3.9rc1
macOS
macOS 26.4
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 192.4 | 13.4 | 16.6 GB | view |
| 4k | 198.9 | 13.5 | 18.1 GB | view |
| 8k | 187.3 | 11.9 | 19.1 GB | view |
| 16k | 163.0 | 11.3 | 20.6 GB | view |
| 32k | 131.1 | 6.6 | 23.6 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 13.4 | 1.00× |
| 2× | 16.5 | 1.23× |
| 4× | 38.5 | 2.87× |