← Back to community benchmarks
Qwen3.6-35B-A3B
Performance
8k
tokens
446.0
PP tok/s
41.6
TG tok/s
18369
TTFT (ms)
20.4
Peak mem (GB)
Hardware
Chip
M4 (10c)
Memory
32 GB
GPU Cores
10
Software
oMLX
v0.3.9.dev2
macOS
macOS 26.0.1
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 406.9 | 46.3 | 19.2 GB | view |
| 4k | 456.5 | 44.1 | 19.9 GB | view |
| 8k | 446.0 | 41.6 | 20.4 GB | current |
| 16k | 407.2 | 35.7 | 21.1 GB | view |
| 32k | 333.0 | 29.8 | 22.6 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 46.3 | 1.00× |
| 2× | 69.6 | 1.50× |
| 4× | 82.3 | 1.78× |
| 8× | 95.0 | 2.05× |