← Back to community benchmarks
Qwen3.6-35B-A3B
Performance
16k
tokens
731.0
PP tok/s
50.6
TG tok/s
22412
TTFT (ms)
37.1
Peak mem (GB)
Hardware
Chip
M2 Max (38c)
Memory
64 GB
GPU Cores
38
Software
oMLX
v0.3.6
macOS
macOS 26.0
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 630.5 | 66.1 | 35.4 GB | view |
| 4k | 755.8 | 62.2 | 36.2 GB | view |
| 8k | 760.6 | 59.3 | 36.5 GB | view |
| 16k | 731.0 | 50.6 | 37.1 GB | current |
| 32k | 664.0 | 39.0 | 38.5 GB | view |
| 64k | 552.7 | 27.0 | 41.2 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 66.1 | 1.00× |
| 2× | 105.4 | 1.59× |
| 4× | 145.1 | 2.20× |
| 8× | 164.6 | 2.49× |