← Back to community benchmarks
Qwen3.6-27B-oQ4-mtp
Performance
32k
tokens
212.4
PP tok/s
17.7
TG tok/s
154259
TTFT (ms)
23.4
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.9
macOS
macOS 26.4.1
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 241.8 | 27.8 | 16.4 GB | view |
| 4k | 244.0 | 26.9 | 17.9 GB | view |
| 8k | 214.6 | 25.5 | 18.9 GB | view |
| 16k | 220.8 | 24.6 | 20.4 GB | view |
| 32k | 212.4 | 17.7 | 23.4 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 27.8 | 1.00× |
| 2× | 34.1 | 1.23× |
| 4× | 40.9 | 1.47× |