I'm running AI on my NAS at 5 tokens per second, and it's surprisingly useful

XDA developers | Aug. 2, 2026, 7:30 a.m. | Original
I ran a local AI model on my NAS with no GPU at about 5 tokens per second. It’s painfully slow, but far more useful than I expected.
I replaced the thermal paste on a five-year-old CPU, and got back performance I didn't know I'd lost Your phone clings to the weak upstairs Wi-Fi, and your router has a setting to make it let go