← Back to community benchmarks
Qwen3.6-35B-A3B
Performance
8k
tokens
3,918
PP tok/s
102.9
TG tok/s
2091
TTFT (ms)
28.6
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.9.dev2
macOS
macOS 26.5
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 2,219 | 109.2 | 27.3 GB | view |
| 4k | 3,684 | 103.6 | 28.1 GB | view |
| 8k | 3,918 | 102.9 | 28.6 GB | current |
| 16k | 3,799 | 99.2 | 29.3 GB | view |
| 32k | 3,216 | 89.0 | 30.8 GB | view |
| 64k | 2,335 | 75.0 | 33.8 GB | view |
| 128k | 1,425 | 55.5 | 39.8 GB | view |
| 195k | 821.8 | 37.5 | 46.7 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 109.2 | 1.00× |
| 2× | 199.8 | 1.83× |
| 4× | 386.1 | 3.54× |