← Back to community benchmarks
Qwen3.6-35B-A3B
Performance
16k
tokens
4,049
PP tok/s
113.0
TG tok/s
4047
TTFT (ms)
21.4
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.5.3
macOS
macOS 26.5
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 3,774 | 125.8 | 19.4 GB | view |
| 4k | 4,797 | 75.3 | 20.2 GB | view |
| 8k | 4,680 | 118.3 | 20.7 GB | view |
| 16k | 4,049 | 113.0 | 21.4 GB | current |
| 32k | 3,496 | 105.3 | 22.9 GB | view |
| 64k | 2,346 | 83.6 | 25.8 GB | view |
| 128k | 1,118 | 59.3 | 31.8 GB | view |
| 195k | 652.6 | 34.4 | 38.7 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 125.8 | 1.00× |
| 2× | 207.2 | 1.65× |
| 4× | 244.6 | 1.94× |
| 8× | 285.8 | 2.27× |