← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
64k
tokens
222.0
PP tok/s
7.7
TG tok/s
295215
TTFT (ms)
23.3
Peak mem (GB)
Hardware
Chip
M4 (10c)
Memory
32 GB
GPU Cores
10
Software
oMLX
v0.2.6
macOS
macOS 26.3.1
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 420.9 | 42.6 | 20.0 GB | view |
| 4k | 423.2 | 39.3 | 20.2 GB | view |
| 8k | 422.7 | 40.3 | 20.4 GB | view |
| 16k | 377.9 | 30.9 | 20.8 GB | view |
| 32k | 305.8 | 22.0 | 21.6 GB | view |
| 64k | 222.0 | 7.7 | 23.3 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 42.6 | 1.00× |
| 2× | 64.7 | 1.52× |
| 4× | 76.9 | 1.81× |