← Back to community benchmarks
Qwen3.5-4B
Performance
4k
tokens
404.8
PP tok/s
37.0
TG tok/s
10119
TTFT (ms)
4.2
Peak mem (GB)
Hardware
Chip
M4 (10c)
Memory
16 GB
GPU Cores
10
Software
oMLX
v0.2.8
macOS
macOS 26.2
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 405.9 | 38.7 | 3.9 GB | view |
| 4k | 404.8 | 37.0 | 4.2 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 38.7 | 1.00× |
| 2× | 73.9 | 1.91× |
| 4× | 93.3 | 2.41× |