← Back to community benchmarks
Qwen3.5-4B-Claude-4.6-Opus-Reasoning-Distilled
Performance
4k
tokens
972.8
PP tok/s
80.3
TG tok/s
4211
TTFT (ms)
3.5
Peak mem (GB)
Hardware
Chip
M3 Max (30c)
Memory
36 GB
GPU Cores
30
Software
oMLX
v0.2.7
macOS
macOS 26.3.1
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 852.4 | 89.3 | 3.4 GB | view |
| 4k | 972.8 | 80.3 | 3.5 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 89.3 | 1.00× |
| 2× | 159.2 | 1.78× |
| 4× | 207.5 | 2.32× |