← Back to community benchmarks
Qwen3.6-27B-oQ4-mtp
Performance
32k
tokens
227.8
PP tok/s
31.7
TG tok/s
143865
TTFT (ms)
23.7
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.9.dev2
macOS
macOS 26.4.1
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 240.7 | 41.4 | 16.7 GB | view |
| 4k | 249.8 | 39.9 | 18.2 GB | view |
| 8k | 247.9 | 37.0 | 19.2 GB | view |
| 16k | 241.2 | 36.1 | 20.7 GB | view |
| 32k | 227.8 | 31.7 | 23.7 GB | current |
| 64k | 202.5 | 25.6 | 29.7 GB | view |
| 128k | 163.5 | 19.3 | 42.1 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 41.4 | 1.00× |
| 2× | 28.4 | 0.69× |
| 4× | 55.9 | 1.35× |