← Back to community benchmarks
Qwen3.5-0.8B
Performance
4k
tokens
1,679
PP tok/s
12.9
TG tok/s
2440
TTFT (ms)
3.4
Peak mem (GB)
Hardware
Chip
M4 (10c)
Memory
16 GB
GPU Cores
10
Software
oMLX
v0.2.8
macOS
macOS 26.3
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,471 | 13.3 | 3.3 GB | view |
| 4k | 1,679 | 12.9 | 3.4 GB | current |
| 8k | 1,611 | 12.4 | 3.6 GB | view |
| 16k | 1,425 | 12.2 | 4.4 GB | view |
| 32k | 1,096 | 11.2 | 5.8 GB | view |
| 64k | 742.4 | 9.6 | 6.7 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 13.3 | 1.00× |
| 2× | 14.3 | 1.08× |
| 4× | 27.2 | 2.05× |
| 8× | 39.2 | 2.95× |