← Back to community benchmarks
Qwen3.6-27B-oQ8-mtp
Performance
8k
tokens
406.3
PP tok/s
18.1
TG tok/s
20162
TTFT (ms)
30.9
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.10
macOS
macOS 26.4
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 297.5 | 18.9 | 28.8 GB | view |
| 4k | 370.4 | 17.1 | 30.3 GB | view |
| 8k | 406.3 | 18.1 | 30.9 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 18.9 | 1.00× |
| 2× | 18.0 | 0.95× |
| 4× | 23.5 | 1.24× |