← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
16k
tokens
2,276
PP tok/s
28.1
TG tok/s
7198
TTFT (ms)
20.8
Peak mem (GB)
Hardware
Chip
M3 Ultra (80c)
Memory
512 GB
GPU Cores
80
Software
oMLX
v0.2.8
macOS
macOS 15.6
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 2,093 | 31.7 | 20.0 GB | view |
| 4k | 2,420 | 29.7 | 20.2 GB | view |
| 8k | 2,437 | 28.5 | 20.4 GB | view |
| 16k | 2,276 | 28.1 | 20.8 GB | current |
| 32k | 1,950 | 24.8 | 21.6 GB | view |
| 64k | 1,450 | 24.6 | 23.3 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 31.7 | 1.00× |
| 2× | 80.3 | 2.53× |
| 4× | 176.7 | 5.57× |
| 8× | 391.8 | 12.36× |