← Back to community benchmarks
Qwen3.6-27B-oQ8-mtp
Performance
16k
tokens
647.1
PP tok/s
15.1
TG tok/s
25320
TTFT (ms)
32.3
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.8
macOS
macOS 26.4.1
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 724.6 | 18.2 | 28.3 GB | view |
| 4k | 746.5 | 15.1 | 29.8 GB | view |
| 8k | 686.2 | 16.2 | 30.8 GB | view |
| 16k | 647.1 | 15.1 | 32.3 GB | current |
| 32k | 599.6 | 10.5 | 35.3 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 18.2 | 1.00× |
| 2× | 30.8 | 1.69× |