← Back to community benchmarks
Qwen3.5-4B
Performance
4k
tokens
398.2
PP tok/s
34.1
TG tok/s
10287
TTFT (ms)
4.6
Peak mem (GB)
Hardware
Chip
M4 (10c)
Memory
24 GB
GPU Cores
10
Software
oMLX
v0.2.8
macOS
macOS 15.2
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 403.6 | 36.7 | 3.9 GB | view |
| 4k | 398.2 | 34.1 | 4.6 GB | current |
| 8k | 386.1 | 32.2 | 5.0 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 36.7 | 1.00× |
| 2× | 69.6 | 1.90× |