← Back to community benchmarks
Qwen3.5-35B-A3B-Claude-4.6-Opus-Reasoning-Distilled-qx86-hi
Performance
8k
tokens
1,588
PP tok/s
80.8
TG tok/s
5158
TTFT (ms)
33.7
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.2.6
macOS
macOS 26.2
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,202 | 87.0 | 33.3 GB | view |
| 4k | 1,576 | 82.3 | 33.5 GB | view |
| 8k | 1,588 | 80.8 | 33.7 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 87.0 | 1.00× |
| 2× | 138.8 | 1.60× |
| 4× | 197.3 | 2.27× |
| 8× | 244.6 | 2.81× |