← Back to community benchmarks
Qwen3.6-27B
Performance
8k
tokens
177.3
PP tok/s
10.2
TG tok/s
46192
TTFT (ms)
35.2
Peak mem (GB)
Hardware
Chip
M3 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.7
macOS
macOS 26.3
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 207.0 | 11.0 | 33.0 GB | view |
| 4k | 176.8 | 10.0 | 34.6 GB | view |
| 8k | 177.3 | 10.2 | 35.2 GB | current |
| 16k | 169.9 | 6.5 | 36.4 GB | view |
| 32k | 151.5 | 7.5 | 38.9 GB | view |
| 64k | 139.3 | 6.7 | 44.0 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 11.0 | 1.00× |
| 2× | 11.5 | 1.05× |
| 4× | 19.0 | 1.73× |
| 8× | 26.5 | 2.41× |