← Back to community benchmarks
Qwen3.6-35B-A3B
Performance
8k
tokens
3,805
PP tok/s
69.5
TG tok/s
2153
TTFT (ms)
66.7
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
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,997 | 73.5 | 65.6 GB | view |
| 4k | 3,546 | 71.8 | 66.4 GB | view |
| 8k | 3,805 | 69.5 | 66.7 GB | current |
| 16k | 3,679 | 65.4 | 67.4 GB | view |
| 32k | 3,220 | 51.1 | 68.7 GB | view |
| 64k | 2,415 | 36.2 | 71.3 GB | view |
| 128k | 1,539 | 23.8 | 76.8 GB | view |
| 195k | 1,082 | 17.5 | 82.5 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 73.5 | 1.00× |
| 2× | 80.8 | 1.10× |
| 4× | 136.9 | 1.86× |
| 8× | 320.0 | 4.35× |