← Back to community benchmarks
Qwen3.6-35B-A3Bmxfp8
Performance
32k
tokens
3,186
PP tok/s
72.3
TG tok/s
10287
TTFT (ms)
37.5
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.5
macOS
macOS 26.4.1
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 2,006 | 97.3 | 34.4 GB | view |
| 4k | 3,483 | 94.2 | 35.2 GB | view |
| 8k | 3,770 | 90.8 | 35.5 GB | view |
| 16k | 3,666 | 83.9 | 36.1 GB | view |
| 32k | 3,186 | 72.3 | 37.5 GB | current |
| 64k | 2,261 | 55.3 | 40.2 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 97.3 | 1.00× |
| 2× | 154.7 | 1.59× |
| 4× | 213.3 | 2.19× |