← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
1k
tokens
2,098
PP tok/s
93.0
TG tok/s
488
TTFT (ms)
20.1
Peak mem (GB)
Hardware
Chip
M3 Ultra (60c)
Memory
256 GB
GPU Cores
60
Software
oMLX
v0.2.6
macOS
macOS 26.3
Context
1,024
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 2,098 | 93.0 | 20.1 GB | current |
| 4k | 2,175 | 92.5 | 20.2 GB | view |
| 8k | 2,103 | 81.7 | 20.4 GB | view |
| 16k | 1,943 | 71.2 | 20.8 GB | view |
| 32k | 1,676 | 61.2 | 21.6 GB | view |
| 64k | 1,295 | 46.6 | 23.3 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 93.0 | 1.00× |
| 2× | 165.8 | 1.78× |
| 4× | 264.7 | 2.85× |
| 8× | 371.4 | 3.99× |