← Back to community benchmarks
Qwen3.6-35B-A3B-UD
Performance
8k
tokens
1,685
PP tok/s
82.5
TG tok/s
4862
TTFT (ms)
21.5
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.8rc1
macOS
macOS 26.3
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,219 | 93.2 | 20.4 GB | view |
| 4k | 1,642 | 88.6 | 21.2 GB | view |
| 8k | 1,685 | 82.5 | 21.5 GB | current |
| 16k | 1,586 | 76.2 | 22.2 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 93.2 | 1.00× |
| 2× | 152.4 | 1.64× |
| 4× | 233.1 | 2.50× |
| 8× | 298.7 | 3.20× |