← Back to community benchmarks
Qwen3.6-27B-oQ6-mtp
User commentQwen3.6-27B-oQ6 model with MTP Optimizations
Performance
16k
tokens
175.8
PP tok/s
18.3
TG tok/s
93200
TTFT (ms)
26.9
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
64 GB
GPU Cores
40
Software
oMLX
v0.3.9
macOS
macOS 26.4.1
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 228.0 | 24.5 | 23.0 GB | view |
| 4k | 193.4 | 21.5 | 24.4 GB | view |
| 8k | 182.2 | 22.8 | 25.4 GB | view |
| 16k | 175.8 | 18.3 | 26.9 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 24.5 | 1.00× |
| 2× | 21.2 | 0.87× |
| 4× | 24.8 | 1.01× |