← Back to community benchmarks
Qwen3.6-35B-A3B
Performance
32k
tokens
1,273
PP tok/s
86.7
TG tok/s
25736
TTFT (ms)
22.7
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
48 GB
GPU Cores
40
Software
oMLX
v0.3.8
macOS
macOS 26.3
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,107 | 109.8 | 19.2 GB | view |
| 4k | 1,652 | 108.0 | 20.0 GB | view |
| 8k | 1,665 | 103.4 | 20.4 GB | view |
| 16k | 1,556 | 95.8 | 21.2 GB | view |
| 32k | 1,273 | 86.7 | 22.7 GB | current |
| 64k | 997.8 | 68.9 | 25.7 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 109.8 | 1.00× |
| 2× | 183.1 | 1.67× |
| 4× | 267.3 | 2.43× |
| 8× | 329.6 | 3.00× |