← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
8k
tokens
718.6
PP tok/s
69.1
TG tok/s
11400
TTFT (ms)
20.4
Peak mem (GB)
Hardware
Chip
M4 Pro (16c)
Memory
48 GB
GPU Cores
16
Software
oMLX
v0.2.9
macOS
macOS 26.3.1
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 727.3 | 83.6 | 20.1 GB | view |
| 4k | 749.1 | 79.0 | 20.2 GB | view |
| 8k | 718.6 | 69.1 | 20.4 GB | current |
| 16k | 665.7 | 53.5 | 20.8 GB | view |
| 32k | 564.6 | 35.6 | 21.6 GB | view |
| 64k | 446.2 | 21.5 | 23.3 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 83.6 | 1.00× |
| 2× | 121.9 | 1.46× |
| 4× | 143.6 | 1.72× |