← Back to community benchmarks
Qwen3.6-35B-A3B-oQ8
Performance
64k
tokens
2,322
PP tok/s
56.5
TG tok/s
28226
TTFT (ms)
41.2
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.6
macOS
macOS 26.4.1
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 2,255 | 100.9 | 35.4 GB | view |
| 4k | 3,698 | 98.3 | 36.2 GB | view |
| 8k | 3,940 | 94.0 | 36.5 GB | view |
| 16k | 3,792 | 84.4 | 37.1 GB | view |
| 32k | 3,241 | 71.2 | 38.5 GB | view |
| 64k | 2,322 | 56.5 | 41.2 GB | current |
| 128k | 1,461 | 40.5 | 46.5 GB | view |
| 195k | 1,026 | 27.0 | 52.2 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 100.9 | 1.00× |
| 2× | 160.8 | 1.59× |
| 4× | 236.7 | 2.35× |
| 8× | 629.0 | 6.23× |