← Back to community benchmarks
Qwen3.5-0.8B
Performance
32k
tokens
3,082
PP tok/s
132.0
TG tok/s
10634
TTFT (ms)
4.1
Peak mem (GB)
Hardware
Chip
M4 Pro (20c)
Memory
64 GB
GPU Cores
20
Software
oMLX
v0.3.6
macOS
macOS 26.4.1
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 2,599 | 219.5 | 1.9 GB | view |
| 4k | 3,733 | 205.9 | 2.1 GB | view |
| 8k | 3,935 | 191.1 | 2.3 GB | view |
| 16k | 3,669 | 164.1 | 2.9 GB | view |
| 32k | 3,082 | 132.0 | 4.1 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 219.5 | 1.00× |
| 2× | 376.6 | 1.72× |
| 4× | 542.0 | 2.47× |
| 8× | 606.1 | 2.76× |