← Back to community benchmarks
Qwen3.6-35B-A3B
Performance
8k
tokens
1,147
PP tok/s
87.2
TG tok/s
7141
TTFT (ms)
19.5
Peak mem (GB)
Hardware
Chip
M3 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.9.dev2
macOS
macOS 26.3
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 852.4 | 96.2 | 18.3 GB | view |
| 4k | 1,190 | 94.0 | 19.0 GB | view |
| 8k | 1,147 | 87.2 | 19.5 GB | current |
| 16k | 1,018 | 73.8 | 20.2 GB | view |
| 32k | 852.2 | 61.0 | 21.7 GB | view |
| 64k | 659.8 | 49.5 | 24.7 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 96.2 | 1.00× |
| 2× | 144.6 | 1.50× |