← Back to community benchmarks
Qwen3.6-27B-oQ4
Performance
16k
tokens
743.4
PP tok/s
28.4
TG tok/s
22038
TTFT (ms)
19.6
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.6
macOS
macOS 26.4.1
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 768.7 | 32.3 | 16.3 GB | view |
| 4k | 887.6 | 31.5 | 17.7 GB | view |
| 8k | 858.7 | 30.5 | 18.4 GB | view |
| 16k | 743.4 | 28.4 | 19.6 GB | current |
| 32k | 623.3 | 21.0 | 22.1 GB | view |
| 64k | 399.6 | 15.6 | 27.1 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 32.3 | 1.00× |
| 2× | 35.9 | 1.11× |
| 4× | 46.7 | 1.45× |