← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
32k
tokens
1,934
PP tok/s
26.4
TG tok/s
16944
TTFT (ms)
21.6
Peak mem (GB)
Hardware
Chip
M3 Ultra (80c)
Memory
512 GB
GPU Cores
80
Software
oMLX
v0.2.8
macOS
macOS 15.6
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 2,122 | 36.3 | 20.0 GB | view |
| 4k | 2,469 | 35.7 | 20.2 GB | view |
| 8k | 2,409 | 31.6 | 20.4 GB | view |
| 16k | 2,261 | 30.1 | 20.8 GB | view |
| 32k | 1,934 | 26.4 | 21.6 GB | current |
| 64k | 1,446 | 22.1 | 23.3 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 36.3 | 1.00× |
| 2× | 64.4 | 1.77× |
| 4× | 134.2 | 3.70× |
| 8× | 208.3 | 5.74× |