← Back to community benchmarks
Qwen3.6-27B-oQ4-fp16-mtp
Performance
8k
tokens
125.3
PP tok/s
20.2
TG tok/s
65380
TTFT (ms)
19.2
Peak mem (GB)
Hardware
Chip
M4 Pro (20c)
Memory
48 GB
GPU Cores
20
Software
oMLX
v0.3.9.dev2
macOS
macOS 26.4.1
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 132.5 | 20.7 | 16.7 GB | view |
| 4k | 134.4 | 22.8 | 18.2 GB | view |
| 8k | 125.3 | 20.2 | 19.2 GB | current |
| 16k | 118.4 | 18.1 | 20.7 GB | view |
| 32k | 105.8 | 16.5 | 23.7 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 20.7 | 1.00× |
| 2× | 23.2 | 1.12× |
| 4× | 47.0 | 2.27× |