← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
64k
tokens
1,603
PP tok/s
46.2
TG tok/s
40875
TTFT (ms)
39.5
Peak mem (GB)
Hardware
Chip
M3 Ultra (80c)
Memory
512 GB
GPU Cores
80
Software
oMLX
v0.2.11
macOS
macOS 26.3.1
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 2,532 | 76.7 | 36.2 GB | view |
| 4k | 2,679 | 74.3 | 36.3 GB | view |
| 8k | 2,609 | 70.3 | 36.5 GB | view |
| 16k | 2,416 | 65.9 | 37.0 GB | view |
| 32k | 2,089 | 58.0 | 37.8 GB | view |
| 64k | 1,603 | 46.2 | 39.5 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 76.7 | 1.00× |
| 2× | 131.8 | 1.72× |
| 4× | 213.0 | 2.78× |