Remix.run Logo
▲ latentsea 5 hours ago

At $500 per month, it's cheaper to just buy GPUs and use local models.

▲tripleee 5 hours ago | parent [-]

Have you looked at the prices of GPUs lately?

▲latentsea 5 hours ago | parent [-]

Yup. I got an R9700 recently for exactly this reason. Figured if I'm going to spend $2400 a year I may as well have something to show for it at the end of it.

That they are expensive and climbing doesn't negate my point if the cost of the subscription over how long you plan to keep it is equally or more expensive than the GPUs. You can put together dual 5060 Ti or 5070 Ti systems to run local LLMs too. You don't need to splurge on a 5090. That's a bad option at this point.

▲tripleee 5 hours ago | parent [-]

What models are you running locally? Are you banking on them improving or do you think they're good enough today? 32GB of VRAM there wouldn't be close to enough to run the best local models.

I've messed around with Qwen3.6-27B but I'm not sure if it could yet even replace Luna for me.