← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
4k
tokens
1,747
PP tok/s
106.9
TG tok/s
2345
TTFT (ms)
20.2
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.2.6
macOS
macOS 26.2
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,669 | 110.2 | 20.0 GB | view |
| 4k | 1,747 | 106.9 | 20.2 GB | current |
| 8k | 1,550 | 97.0 | 20.4 GB | view |
| 16k | 1,303 | 88.1 | 20.8 GB | view |
| 32k | 1,066 | 57.5 | 21.6 GB | view |
| 64k | 761.0 | 20.6 | 23.3 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 110.2 | 1.00× |
| 2× | 158.7 | 1.44× |
| 4× | 147.2 | 1.34× |
| 8× | 150.3 | 1.36× |