← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
8k
tokens
439.8
PP tok/s
44.7
TG tok/s
18626
TTFT (ms)
28.5
Peak mem (GB)
Hardware
Chip
M1 Max (32c)
Memory
64 GB
GPU Cores
32
Software
oMLX
v0.2.7
macOS
macOS 26.3.1
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 506.5 | 50.7 | 28.1 GB | view |
| 4k | 487.3 | 48.9 | 28.3 GB | view |
| 8k | 439.8 | 44.7 | 28.5 GB | current |
| 16k | 401.3 | 33.9 | 28.9 GB | view |
| 32k | 379.9 | 30.2 | 29.7 GB | view |
| 64k | 325.6 | 20.3 | 31.4 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 50.7 | 1.00× |
| 2× | 63.4 | 1.25× |
| 4× | 75.8 | 1.50× |
| 8× | 99.1 | 1.95× |