← Back to community benchmarks
Qwen3.6-27B-oQ8-fp16-mtp
Performance
8k
tokens
71.7
PP tok/s
9.4
TG tok/s
114237
TTFT (ms)
31.3
Peak mem (GB)
Hardware
Chip
M1 Max (32c)
Memory
64 GB
GPU Cores
32
Software
oMLX
v0.3.9
macOS
macOS 26.5
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 92.5 | 11.4 | 28.8 GB | view |
| 4k | 76.6 | 9.5 | 30.3 GB | view |
| 8k | 71.7 | 9.4 | 31.3 GB | current |
| 16k | 73.8 | 10.2 | 32.8 GB | view |
| 32k | 71.0 | 9.5 | 35.8 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 11.4 | 1.00× |
| 2× | 12.3 | 1.08× |
| 4× | 14.7 | 1.29× |
| 8× | 21.1 | 1.85× |