← Back to community benchmarks
Qwen3.6-35B-A3B
Performance
16k
tokens
1,981
PP tok/s
72.9
TG tok/s
8271
TTFT (ms)
21.0
Peak mem (GB)
Hardware
Chip
M5 Pro (20c)
Memory
48 GB
GPU Cores
20
Software
oMLX
v0.3.5
macOS
macOS 26.5
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,386 | 93.0 | 19.3 GB | view |
| 4k | 2,084 | 91.5 | 20.0 GB | view |
| 8k | 2,081 | 80.6 | 20.4 GB | view |
| 16k | 1,981 | 72.9 | 21.0 GB | current |
| 32k | 1,609 | 56.6 | 22.4 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 93.0 | 1.00× |
| 2× | 134.1 | 1.44× |
| 4× | 268.1 | 2.88× |
| 8× | 540.8 | 5.82× |