← Back to community benchmarks
Qwen3.6-27B-oQ4-mtp
Performance
1k
tokens
741.6
PP tok/s
41.5
TG tok/s
1381
TTFT (ms)
16.8
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.9.dev1
macOS
macOS 26.4.1
Context
1,024
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 741.6 | 41.5 | 16.8 GB | current |
| 4k | 883.8 | 46.6 | 18.2 GB | view |
| 8k | 830.4 | 44.2 | 19.2 GB | view |
| 16k | 745.4 | 41.7 | 20.7 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 41.5 | 1.00× |
| 2× | 55.7 | 1.34× |
| 4× | 68.3 | 1.65× |