← Back to community benchmarks
Qwen3.6-35B-A3B-4bit-DWQ
Performance
32k
tokens
2,223
PP tok/s
68.7
TG tok/s
14738
TTFT (ms)
23.4
Peak mem (GB)
Hardware
Chip
M3 Ultra (80c)
Memory
256 GB
GPU Cores
80
Software
oMLX
v0.3.6
macOS
macOS 26.4.1
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,760 | 96.4 | 20.3 GB | view |
| 4k | 2,605 | 89.4 | 21.1 GB | view |
| 8k | 2,695 | 87.0 | 21.5 GB | view |
| 16k | 2,568 | 80.2 | 22.1 GB | view |
| 32k | 2,223 | 68.7 | 23.4 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 96.4 | 1.00× |
| 2× | 163.5 | 1.70× |