← Back to community benchmarks
Qwen3.6-27B-oQ8-mtp
Performance
8k
tokens
247.2
PP tok/s
25.9
TG tok/s
33144
TTFT (ms)
31.3
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.9
macOS
macOS 26.4
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 239.1 | 26.9 | 28.8 GB | view |
| 4k | 248.8 | 26.8 | 30.3 GB | view |
| 8k | 247.2 | 25.9 | 31.3 GB | current |
| 16k | 240.8 | 24.4 | 32.8 GB | view |
| 32k | 227.4 | 22.7 | 35.8 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 26.9 | 1.00× |
| 2× | 31.6 | 1.17× |
| 4× | 45.2 | 1.68× |