← Back to community benchmarks

Qwen3.5-0.8B

M4 (10c) · 16 GB · unknown · 2026-03-11
Performance
4k
tokens
1,438
PP tok/s
11.7
TG tok/s
2848
TTFT (ms)
3.4
Peak mem (GB)
Hardware
Chip M4 (10c)
Memory 16 GB
GPU Cores 10
Software
oMLX v0.2.7
macOS macOS 15.2
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,087 11.9 3.3 GB view
4k 1,438 11.7 3.4 GB current
Batching Results
Batch Size TG tok/s Speedup
11.9 1.00×
12.6 1.06×
34.9 2.93×