← Back to community benchmarks
Qwen3.6-27B
Performance
32k
tokens
113.8
PP tok/s
8.0
TG tok/s
287870
TTFT (ms)
35.3
Peak mem (GB)
Hardware
Chip
M4 Pro (20c)
Memory
48 GB
GPU Cores
20
Software
oMLX
v0.4.4rc1
macOS
macOS 26.5
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 118.5 | 8.7 | 28.1 GB | view |
| 4k | 123.4 | 8.6 | 29.4 GB | view |
| 8k | 123.9 | 8.5 | 30.8 GB | view |
| 16k | 120.6 | 8.3 | 32.3 GB | view |
| 32k | 113.8 | 8.0 | 35.3 GB | current |
| 64k | 71.4 | 7.3 | 37.0 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 8.7 | 1.00× |
| 2× | 17.0 | 1.95× |
| 4× | 23.7 | 2.72× |
| 8× | 24.5 | 2.82× |