Remix.run Logo
▲ paxy 3 hours ago

What are you disagreeing with exactly? Qwen 3.8 flash costs a few cents per million tokens from cloud providers. You can get effectively unlimited usage on a $6/month plan. You were okay with buying a $5000+ Mac Studio for it because you have the money, but you are in the tiny minority.

▲nsbk 2 hours ago | parent [-]

I disagree with your statement saying there's no difference between self-hosting local AI and self-hosting regular software. Regular services (which I guess is what you're referring to) require uptime, to be accesible from the Internet, backups, security patches, the list goes on. The main argument against self-hosting has never been cost, it is a complex task that requires knowledge and dedication. On the other hand, you can self-host on a Raspberry Pi or ancient laptop for close to nothing.

A local AI can be as simple as a program that runs in your computer [0]. Any machine with 12GB+ of VRAM can run the same model I'm using. I has a chat interface, can see images, control your browser, generate artifacts, work on projects and workspaces. You name it.

[0] https://stratallm.org