← Back to community benchmarks
Qwen3.5-4B
Performance
32k
tokens
315.2
PP tok/s
23.2
TG tok/s
103945
TTFT (ms)
5.9
Peak mem (GB)
Hardware
Chip
M4 (10c)
Memory
16 GB
GPU Cores
10
Software
oMLX
v0.2.10
macOS
macOS 26.3.1
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 406.6 | 38.6 | 3.9 GB | view |
| 4k | 404.7 | 36.8 | 4.2 GB | view |
| 8k | 389.4 | 33.3 | 4.6 GB | view |
| 16k | 365.1 | 29.1 | 4.8 GB | view |
| 32k | 315.2 | 23.2 | 5.9 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 38.6 | 1.00× |
| 2× | 73.7 | 1.91× |
| 4× | 86.1 | 2.23× |