← Back to community benchmarks
Qwen3.6-27B-oQ4-mtp
Performance
32k
tokens
677.3
PP tok/s
26.9
TG tok/s
48379
TTFT (ms)
23.4
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.9
macOS
macOS 26.5
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 645.7 | 32.0 | 16.5 GB | view |
| 4k | 741.1 | 31.4 | 17.9 GB | view |
| 8k | 761.1 | 30.8 | 18.9 GB | view |
| 16k | 748.5 | 29.5 | 20.4 GB | view |
| 32k | 677.3 | 26.9 | 23.4 GB | current |
| 64k | 543.9 | 22.6 | 29.5 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 32.0 | 1.00× |
| 2× | 48.6 | 1.52× |
| 4× | 59.6 | 1.86× |
| 8× | 65.6 | 2.05× |