← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
8k
tokens
1,382
PP tok/s
91.9
TG tok/s
5930
TTFT (ms)
20.4
Peak mem (GB)
Hardware
Chip
M4 Max (32c)
Memory
36 GB
GPU Cores
32
Software
oMLX
v0.2.6
macOS
macOS 26.3
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,411 | 99.7 | 20.0 GB | view |
| 4k | 1,431 | 97.4 | 20.2 GB | view |
| 8k | 1,382 | 91.9 | 20.4 GB | current |
| 16k | 1,273 | 77.7 | 20.8 GB | view |
| 32k | 1,095 | 59.9 | 21.6 GB | view |
| 64k | 845.0 | 40.7 | 23.3 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 99.7 | 1.00× |
| 2× | 167.4 | 1.68× |
| 4× | 230.7 | 2.31× |
| 8× | 297.9 | 2.99× |