← Back to community benchmarks
Qwen3.6-27B
Performance
64k
tokens
437.2
PP tok/s
20.9
TG tok/s
149905
TTFT (ms)
30.2
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.5.1
macOS
macOS 26.5
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 525.3 | 27.1 | 15.8 GB | view |
| 4k | 543.6 | 26.1 | 17.3 GB | view |
| 8k | 605.2 | 30.0 | 18.3 GB | view |
| 16k | 623.8 | 26.6 | 20.4 GB | view |
| 32k | 511.9 | 21.3 | 23.6 GB | view |
| 64k | 437.2 | 20.9 | 30.2 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 27.1 | 1.00× |
| 2× | 47.2 | 1.74× |
| 4× | 69.0 | 2.55× |
| 8× | 83.0 | 3.06× |