← Back to community benchmarks
Qwen3.6-35B-A3B-oQ6
Performance
32k
tokens
1,694
PP tok/s
48.6
TG tok/s
19345
TTFT (ms)
30.8
Peak mem (GB)
Hardware
Chip
M5 Pro (20c)
Memory
64 GB
GPU Cores
20
Software
oMLX
v0.3.6
macOS
macOS 26.4.1
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,179 | 70.5 | 27.7 GB | view |
| 4k | 2,034 | 68.1 | 28.5 GB | view |
| 8k | 2,104 | 65.5 | 28.9 GB | view |
| 16k | 1,983 | 59.4 | 29.5 GB | view |
| 32k | 1,694 | 48.6 | 30.8 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 70.5 | 1.00× |
| 2× | 109.0 | 1.55× |
| 4× | 148.1 | 2.10× |