← Back to community benchmarks
Qwen3.6-35B-A3B
Performance
64k
tokens
2,310
PP tok/s
70.7
TG tok/s
28370
TTFT (ms)
42.2
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.5.1
macOS
macOS 26.5.2
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 2,321 | 94.7 | 35.4 GB | view |
| 4k | 4,125 | 92.9 | 36.2 GB | view |
| 8k | 4,322 | 91.0 | 36.7 GB | view |
| 16k | 3,832 | 87.0 | 37.6 GB | view |
| 32k | 3,365 | 81.6 | 39.4 GB | view |
| 64k | 2,310 | 70.7 | 42.2 GB | current |
| 128k | 1,446 | 55.7 | 47.8 GB | view |
| 195k | 930.2 | 44.1 | 55.5 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 94.7 | 1.00× |
| 2× | 154.4 | 1.63× |
| 4× | 242.2 | 2.56× |
| 8× | 324.8 | 3.43× |