← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
16k
tokens
797.6
PP tok/s
60.7
TG tok/s
20541
TTFT (ms)
20.8
Peak mem (GB)
Hardware
Chip
M4 Pro (20c)
Memory
48 GB
GPU Cores
20
Software
oMLX
v0.2.7
macOS
macOS 26.3
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 890.5 | 86.2 | 20.1 GB | view |
| 4k | 908.7 | 79.5 | 20.2 GB | view |
| 8k | 877.6 | 74.9 | 20.4 GB | view |
| 16k | 797.6 | 60.7 | 20.8 GB | current |
| 32k | 686.8 | 51.8 | 21.6 GB | view |
| 64k | 531.7 | 35.1 | 23.3 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 86.2 | 1.00× |
| 2× | 130.0 | 1.51× |
| 4× | 164.1 | 1.90× |
| 8× | 198.1 | 2.30× |