Decode · tok/s
41.4
Prefill
—
tok/100W
103.5
Here's a microwave built entirely on a MacBook Pro M5 Pro (64 GB) with Qwen 3.8 Flash Next, quantized to ~2.6 bits, served by Sushi and driven by Pi. ... Decode 41.4 tok/s
First-person MacBook Pro M5 Pro 64GB. Rank is the quoted 41.4 decode during a 17 h multi-session agent run. Quant is ~2.6 bits; runtime is Sushi. Peak prompt 110K, only 0.35 GiB swapped.
HF