← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
64k
tokens
372.2
PP tok/s
19.2
TG tok/s
176055
TTFT (ms)
23.3
Peak mem (GB)
Hardware
Chip
M1 Max (32c)
Memory
64 GB
GPU Cores
32
Software
oMLX
v0.2.7
macOS
macOS 26.3.1
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 537.5 | 59.6 | 20.1 GB | view |
| 4k | 563.6 | 57.7 | 20.2 GB | view |
| 8k | 554.2 | 51.4 | 20.4 GB | view |
| 16k | 527.4 | 42.0 | 20.8 GB | view |
| 32k | 476.4 | 31.0 | 21.6 GB | view |
| 64k | 372.2 | 19.2 | 23.3 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 59.6 | 1.00× |
| 2× | 90.3 | 1.52× |
| 4× | 116.1 | 1.95× |
| 8× | 128.9 | 2.16× |