← Back to community benchmarks
qwen3.6-35B-A3B-OptiQ
Performance
64k
tokens
2,429
PP tok/s
79.0
TG tok/s
26987
TTFT (ms)
28.2
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.4.2rc1
macOS
macOS 26.5.1
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 2,127 | 111.1 | 21.7 GB | view |
| 4k | 3,889 | 104.5 | 22.5 GB | view |
| 8k | 4,176 | 104.7 | 23.0 GB | view |
| 16k | 3,960 | 101.1 | 23.7 GB | view |
| 32k | 3,327 | 87.6 | 25.2 GB | view |
| 64k | 2,429 | 79.0 | 28.2 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 111.1 | 1.00× |
| 2× | 142.8 | 1.29× |
| 4× | 241.2 | 2.17× |
| 8× | 334.6 | 3.01× |