← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
4k
tokens
1,997
PP tok/s
64.2
TG tok/s
2051
TTFT (ms)
36.3
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
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,820 | 66.2 | 36.2 GB | view |
| 4k | 1,997 | 64.2 | 36.3 GB | current |
| 8k | 1,967 | 61.7 | 36.5 GB | view |
| 16k | 1,845 | 56.0 | 37.0 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 66.2 | 1.00× |
| 2× | 104.5 | 1.58× |
| 4× | 146.4 | 2.21× |
| 8× | 194.6 | 2.94× |