← Back to community benchmarks
Qwen3.6-27B-oQ8-mtp
User commentQwen3.6-27B-oQ8 model without MTP Optimizations
Performance
16k
tokens
167.3
PP tok/s
10.6
TG tok/s
97937
TTFT (ms)
32.3
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 | 233.6 | 16.2 | 28.3 GB | view |
| 4k | 200.6 | 15.0 | 29.8 GB | view |
| 8k | 184.7 | 15.0 | 30.8 GB | view |
| 16k | 167.3 | 10.6 | 32.3 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 16.2 | 1.00× |
| 2× | 19.7 | 1.22× |
| 4× | 26.5 | 1.64× |