← Back to community benchmarks
Qwen3.6-27B-oQ4-mtp
Performance
8k
tokens
394.8
PP tok/s
40.7
TG tok/s
20749
TTFT (ms)
19.2
Peak mem (GB)
Hardware
Chip
M3 Ultra (80c)
Memory
512 GB
GPU Cores
80
Software
oMLX
v0.3.9.dev2
macOS
macOS 26.3.1
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 357.6 | 40.6 | 16.7 GB | view |
| 4k | 393.4 | 42.4 | 18.2 GB | view |
| 8k | 394.8 | 40.7 | 19.2 GB | current |
| 16k | 385.6 | 38.9 | 20.7 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 40.6 | 1.00× |
| 2× | 60.1 | 1.48× |
| 4× | 80.1 | 1.97× |
| 8× | 85.8 | 2.11× |