← Back to community benchmarks
Qwen3.6-35B-A3B
Performance
32k
tokens
1,025
PP tok/s
61.4
TG tok/s
31982
TTFT (ms)
22.4
Peak mem (GB)
Hardware
Chip
M3 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.8rc1
macOS
macOS 26.3.1
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 963.8 | 90.1 | 19.3 GB | view |
| 4k | 1,426 | 86.0 | 20.0 GB | view |
| 8k | 1,455 | 80.7 | 20.4 GB | view |
| 16k | 1,362 | 72.3 | 21.0 GB | view |
| 32k | 1,025 | 61.4 | 22.4 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 90.1 | 1.00× |
| 2× | 149.4 | 1.66× |
| 4× | 202.4 | 2.25× |
| 8× | 251.2 | 2.79× |