← Back to community benchmarks
Qwen3.6-35B-A3B-oQ6
User commentM5 max 128gb, Battery - High Power mode
Performance
4k
tokens
3,638
PP tok/s
102.5
TG tok/s
1126
TTFT (ms)
28.5
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
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 2,081 | 104.9 | 27.7 GB | view |
| 4k | 3,638 | 102.5 | 28.5 GB | current |
| 8k | 3,908 | 97.7 | 28.9 GB | view |
| 16k | 3,770 | 91.1 | 29.5 GB | view |
| 32k | 3,129 | 79.6 | 30.8 GB | view |
| 64k | 2,222 | 61.4 | 33.5 GB | view |
| 128k | 1,461 | 41.9 | 38.9 GB | view |
| 195k | 1,116 | 31.1 | 44.5 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 104.9 | 1.00× |
| 2× | 165.8 | 1.58× |
| 4× | 249.5 | 2.38× |
| 8× | 332.7 | 3.17× |