← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
32k
tokens
510.2
PP tok/s
39.8
TG tok/s
64228
TTFT (ms)
37.8
Peak mem (GB)
Hardware
Chip
M2 Max (38c)
Memory
96 GB
GPU Cores
38
Software
oMLX
v0.2.13
macOS
macOS 26.3.1
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 726.5 | 64.4 | 36.2 GB | view |
| 4k | 687.2 | 57.8 | 36.3 GB | view |
| 8k | 609.4 | 52.4 | 36.5 GB | view |
| 16k | 575.2 | 48.6 | 37.0 GB | view |
| 32k | 510.2 | 39.8 | 37.8 GB | current |
| 64k | 414.5 | 29.5 | 39.5 GB | view |
| 128k | 283.2 | 13.6 | 42.8 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 64.4 | 1.00× |
| 2× | 48.7 | 0.76× |