← Back to community benchmarks
Qwen3.5-4B
Performance
32k
tokens
320.7
PP tok/s
21.7
TG tok/s
102167
TTFT (ms)
5.9
Peak mem (GB)
Hardware
Chip
M4 (10c)
Memory
16 GB
GPU Cores
10
Software
oMLX
v0.2.7
macOS
macOS 15.5
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 406.4 | 37.4 | 3.9 GB | view |
| 4k | 402.8 | 34.7 | 4.2 GB | view |
| 8k | 392.0 | 32.7 | 4.6 GB | view |
| 16k | 366.7 | 27.2 | 4.8 GB | view |
| 32k | 320.7 | 21.7 | 5.9 GB | current |
| 64k | 254.5 | 14.9 | 8.0 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 37.4 | 1.00× |
| 2× | 70.9 | 1.90× |
| 4× | 92.3 | 2.47× |
| 8× | 94.0 | 2.51× |