← Back to community benchmarks
Qwen3.6-35B-A3B-oQ6
Performance
16k
tokens
1,534
PP tok/s
70.5
TG tok/s
10682
TTFT (ms)
29.4
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.7
macOS
macOS 26.4.1
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,226 | 90.5 | 27.6 GB | view |
| 4k | 1,651 | 86.1 | 28.4 GB | view |
| 8k | 1,656 | 80.2 | 28.8 GB | view |
| 16k | 1,534 | 70.5 | 29.4 GB | current |
| 32k | 1,218 | 55.6 | 30.8 GB | view |
| 64k | 943.6 | 38.4 | 33.4 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 90.5 | 1.00× |
| 2× | 139.4 | 1.54× |
| 4× | 195.4 | 2.16× |
| 8× | 237.4 | 2.62× |