← Back to community benchmarks
Qwen3.6-27B-oQ4-mtp
Performance
32k
tokens
183.6
PP tok/s
18.4
TG tok/s
178516
TTFT (ms)
23.4
Peak mem (GB)
Hardware
Chip
M3 Max (40c)
Memory
64 GB
GPU Cores
40
Software
oMLX
v0.3.9.dev2
macOS
macOS 26.5
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 207.6 | 22.5 | 16.4 GB | view |
| 4k | 205.0 | 21.7 | 17.9 GB | view |
| 8k | 188.9 | 20.2 | 18.9 GB | view |
| 16k | 186.8 | 19.3 | 20.4 GB | view |
| 32k | 183.6 | 18.4 | 23.4 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 22.5 | 1.00× |
| 2× | 35.7 | 1.59× |
| 4× | 42.8 | 1.90× |