← Back to community benchmarks
Qwen3.5-4B
Performance
1k
tokens
332.4
PP tok/s
33.1
TG tok/s
3081
TTFT (ms)
3.9
Peak mem (GB)
Hardware
Chip
M3 (10c)
Memory
16 GB
GPU Cores
10
Software
oMLX
v0.2.6
macOS
macOS 26.2
Context
1,024
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 332.4 | 33.1 | 3.9 GB | current |
| 4k | 332.6 | 31.1 | 4.2 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 33.1 | 1.00× |
| 2× | 64.1 | 1.94× |
| 4× | 77.7 | 2.35× |
| 8× | 79.1 | 2.39× |