← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
1k
tokens
999.7
PP tok/s
60.4
TG tok/s
1024
TTFT (ms)
20.0
Peak mem (GB)
Hardware
Chip
M1 Ultra (64c)
Memory
64 GB
GPU Cores
64
Software
oMLX
v0.2.7
macOS
macOS 15.7.3
Context
1,024
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 999.7 | 60.4 | 20.0 GB | current |
| 4k | 1,087 | 58.0 | 20.2 GB | view |
| 8k | 1,073 | 55.3 | 20.4 GB | view |
| 16k | 1,024 | 44.9 | 20.8 GB | view |
| 32k | 925.1 | 32.4 | 21.6 GB | view |
| 64k | 692.1 | 21.0 | 23.3 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 60.4 | 1.00× |
| 2× | 102.4 | 1.70× |
| 4× | 165.6 | 2.74× |
| 8× | 228.7 | 3.79× |