← Back to community benchmarks
Qwen3.5-4B
Performance
32k
tokens
651.6
PP tok/s
37.9
TG tok/s
50286
TTFT (ms)
6.9
Peak mem (GB)
Hardware
Chip
M4 Pro (20c)
Memory
24 GB
GPU Cores
20
Software
oMLX
v0.2.7
macOS
macOS 15.7.4
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 798.2 | 83.9 | 4.0 GB | view |
| 4k | 803.1 | 78.6 | 4.2 GB | view |
| 8k | 780.2 | 72.9 | 4.6 GB | view |
| 16k | 734.7 | 63.8 | 5.3 GB | view |
| 32k | 651.6 | 37.9 | 6.9 GB | current |
| 64k | 531.3 | 22.9 | 8.6 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 83.9 | 1.00× |
| 2× | 150.3 | 1.79× |
| 4× | 174.1 | 2.08× |
| 8× | 186.3 | 2.22× |