← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
64k
tokens
1,298
PP tok/s
36.3
TG tok/s
50496
TTFT (ms)
42.1
Peak mem (GB)
Hardware
Chip
M5 Pro (20c)
Memory
64 GB
GPU Cores
20
Software
oMLX
v0.2.20
macOS
macOS 26.4
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,730 | 67.2 | 36.2 GB | view |
| 4k | 2,007 | 64.9 | 37.1 GB | view |
| 32k | 1,736 | 48.5 | 39.5 GB | view |
| 64k | 1,298 | 36.3 | 42.1 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 67.2 | 1.00× |
| 2× | 103.4 | 1.54× |
| 4× | 138.0 | 2.05× |