← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
32k
tokens
1,582
PP tok/s
40.2
TG tok/s
20716
TTFT (ms)
37.8
Peak mem (GB)
Hardware
Chip
M5 Pro (20c)
Memory
64 GB
GPU Cores
20
Software
oMLX
v0.2.10
macOS
macOS 26.3.1
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,380 | 59.3 | 36.2 GB | view |
| 4k | 1,620 | 58.8 | 36.3 GB | view |
| 8k | 1,777 | 50.9 | 36.5 GB | view |
| 16k | 1,772 | 55.3 | 37.0 GB | view |
| 32k | 1,582 | 40.2 | 37.8 GB | current |
| 64k | 1,188 | 28.9 | 39.5 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 59.3 | 1.00× |
| 2× | 105.7 | 1.78× |
| 4× | 144.9 | 2.44× |
| 8× | 201.3 | 3.39× |