← Back to community benchmarks
Qwen3.6-35B-A3B
Performance
195k
tokens
606.6
PP tok/s
43.7
TG tok/s
329691
TTFT (ms)
32.8
Peak mem (GB)
Hardware
Chip
M3 Ultra (60c)
Memory
256 GB
GPU Cores
60
Software
oMLX
v0.5.2.dev1
macOS
macOS 27.0
Context
200,000
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,987 | 83.9 | 27.4 GB | view |
| 4k | 2,251 | 84.4 | 28.1 GB | view |
| 8k | 2,162 | 82.7 | 28.3 GB | view |
| 16k | 1,939 | 77.4 | 28.4 GB | view |
| 32k | 1,621 | 72.7 | 28.8 GB | view |
| 64k | 1,221 | 61.8 | 29.5 GB | view |
| 128k | 818.6 | 51.1 | 31.1 GB | view |
| 195k | 606.6 | 43.7 | 32.8 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 83.9 | 1.00× |
| 2× | 149.7 | 1.78× |