← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
32k
tokens
1,358
PP tok/s
49.6
TG tok/s
24123
TTFT (ms)
21.6
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.2.7
macOS
macOS 26.2
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,568 | 113.7 | 20.0 GB | view |
| 4k | 1,703 | 111.3 | 20.2 GB | view |
| 16k | 1,580 | 92.0 | 20.8 GB | view |
| 32k | 1,358 | 49.6 | 21.6 GB | current |
| 64k | 1,049 | 28.1 | 23.3 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 113.7 | 1.00× |
| 2× | 182.0 | 1.60× |
| 4× | 267.3 | 2.35× |
| 8× | 359.4 | 3.16× |