← Back to community benchmarks
Qwen3.5-4B-OptiQ
Performance
4k
tokens
396.3
PP tok/s
31.2
TG tok/s
10335
TTFT (ms)
4.1
Peak mem (GB)
Hardware
Chip
M4 (10c)
Memory
16 GB
GPU Cores
10
Software
oMLX
v0.2.9
macOS
macOS 26.3.1
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 377.9 | 32.3 | 3.8 GB | view |
| 4k | 396.3 | 31.2 | 4.1 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 32.3 | 1.00× |
| 2× | 62.5 | 1.93× |
| 4× | 86.0 | 2.66× |