← Back to community benchmarks
Qwen3.6-35B-A3B
Performance
8k
tokens
1,389
PP tok/s
80.4
TG tok/s
5898
TTFT (ms)
20.5
Peak mem (GB)
Hardware
Chip
M3 Max (40c)
Memory
64 GB
GPU Cores
40
Software
oMLX
v0.3.0rc1
macOS
macOS 15.7.3
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 994.2 | 78.7 | 19.3 GB | view |
| 4k | 1,343 | 79.9 | 20.2 GB | view |
| 8k | 1,389 | 80.4 | 20.5 GB | current |
| 16k | 1,188 | 62.4 | 21.2 GB | view |
| 32k | 960.7 | 47.2 | 22.5 GB | view |
| 64k | 759.8 | 27.8 | 25.2 GB | view |
| 128k | 366.7 | 13.8 | 30.5 GB | view |
| 195k | 273.7 | 9.8 | 36.2 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 78.7 | 1.00× |
| 2× | 125.9 | 1.60× |
| 4× | 160.5 | 2.04× |