← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
32k
tokens
1,089
PP tok/s
49.6
TG tok/s
30101
TTFT (ms)
21.6
Peak mem (GB)
Hardware
Chip
M3 Max (40c)
Memory
48 GB
GPU Cores
40
Software
oMLX
v0.2.8
macOS
macOS 26.3.1
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,474 | 93.6 | 20.0 GB | view |
| 4k | 1,507 | 89.0 | 20.2 GB | view |
| 8k | 1,452 | 82.5 | 20.4 GB | view |
| 16k | 1,343 | 71.3 | 20.8 GB | view |
| 32k | 1,089 | 49.6 | 21.6 GB | current |
| 64k | 881.7 | 32.2 | 23.3 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 93.6 | 1.00× |
| 2× | 149.1 | 1.59× |
| 4× | 230.0 | 2.46× |
| 8× | 282.4 | 3.02× |