← Back to community benchmarks
Qwen3.6-27B-OptiQ
Performance
64k
tokens
168.1
PP tok/s
10.5
TG tok/s
389819
TTFT (ms)
29.9
Peak mem (GB)
Hardware
Chip
M5 Pro (20c)
Memory
48 GB
GPU Cores
20
Software
oMLX
v0.5.2.dev1
macOS
macOS 26.5.2
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 457.2 | 15.1 | 19.2 GB | view |
| 4k | 467.0 | 14.9 | 20.6 GB | view |
| 8k | 459.3 | 14.7 | 21.7 GB | view |
| 16k | 417.0 | 14.2 | 23.2 GB | view |
| 32k | 380.0 | 13.3 | 26.2 GB | view |
| 64k | 168.1 | 10.5 | 29.9 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 15.1 | 1.00× |
| 2× | 24.9 | 1.65× |
| 4× | 41.4 | 2.74× |