← Back to community benchmarks
Qwen3.6-35B-A3B-oQ6
Performance
32k
tokens
3,296
PP tok/s
88.7
TG tok/s
9941
TTFT (ms)
31.1
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.8
macOS
macOS 26.4.1
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 2,218 | 107.4 | 27.7 GB | view |
| 4k | 3,728 | 104.1 | 28.4 GB | view |
| 8k | 3,956 | 102.5 | 28.8 GB | view |
| 16k | 3,842 | 96.8 | 29.6 GB | view |
| 32k | 3,296 | 88.7 | 31.1 GB | current |
| 64k | 2,413 | 74.6 | 34.1 GB | view |
| 128k | 1,515 | 59.0 | 40.1 GB | view |
| 195k | 1,042 | 45.5 | 47.0 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 107.4 | 1.00× |
| 2× | 154.0 | 1.43× |
| 4× | 233.0 | 2.17× |
| 8× | 469.1 | 4.37× |