← Back to community benchmarks
Qwen3.6-27B-oQ8-mtp
User commentQwen3.6-27B-oQ8 model with MTP Optimizations
Performance
16k
tokens
183.8
PP tok/s
18.7
TG tok/s
89130
TTFT (ms)
32.8
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
64 GB
GPU Cores
40
Software
oMLX
v0.3.9rc1
macOS
macOS 26.4.1
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 235.9 | 24.7 | 28.8 GB | view |
| 4k | 198.8 | 22.8 | 30.3 GB | view |
| 8k | 192.4 | 22.1 | 31.3 GB | view |
| 16k | 183.8 | 18.7 | 32.8 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 24.7 | 1.00× |
| 2× | 20.1 | 0.81× |
| 4× | 28.1 | 1.14× |