← Back to community benchmarks
Qwen3.6-35B-A3B
Performance
32k
tokens
3,359
PP tok/s
81.7
TG tok/s
9755
TTFT (ms)
39.3
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
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 3,443 | 94.5 | 35.4 GB | view |
| 4k | 4,497 | 93.1 | 36.2 GB | view |
| 8k | 4,392 | 91.2 | 36.7 GB | view |
| 16k | 4,043 | 87.3 | 37.4 GB | view |
| 32k | 3,359 | 81.7 | 39.3 GB | current |
| 64k | 2,426 | 70.5 | 42.0 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 94.5 | 1.00× |
| 2× | 159.5 | 1.69× |
| 4× | 241.6 | 2.56× |