← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
1k
tokens
1,251
PP tok/s
91.1
TG tok/s
819
TTFT (ms)
19.3
Peak mem (GB)
Hardware
Chip
M5 Pro (20c)
Memory
64 GB
GPU Cores
20
Software
oMLX
v0.3.4
macOS
macOS 26.4
Context
1,024
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,251 | 91.1 | 19.3 GB | current |
| 4k | 1,994 | 88.5 | 20.0 GB | view |
| 8k | 2,082 | 81.5 | 20.4 GB | view |
| 16k | 1,977 | 72.9 | 21.0 GB | view |
| 32k | 1,613 | 59.1 | 22.3 GB | view |
| 64k | 1,129 | 44.0 | 25.0 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 91.1 | 1.00× |
| 2× | 141.8 | 1.56× |
| 4× | 188.6 | 2.07× |