llmdeal is on hold. There is no funding, and without it nothing can move forward. Here is the honest version.
· llmdeal.me
I built llmdeal myself. The idea is sound. One API key, routing across providers, so one vendor's bad week does not become yours. The code works and people tried it.
It has been running on free provider tiers, though. That was always just a starting point. Free tiers rate-limit you, change without warning, and are not built to carry a real product. You cannot put a paying customer on that. Making llmdeal real costs money: proper backends, redundancy, the stronger models. Money means revenue or funding.
llmdeal was never meant to run expensive models for everything. It was the opposite. Cheap and open-weight models do the everyday work, which is most of it. A GPU and the frontier models sit behind that, for the hard requests nothing cheaper can handle. Most traffic never needs them.
The GPU is still available. It will not be deployed without upfront payment. That tier gets paid for before it runs, never fronted. If someone needs that capacity, they pay first and then it spins up.
There is no funding, and not enough revenue to pay for the infrastructure llmdeal needs to be worth paying for. I could keep it running on free tiers and hope. That would mean selling something I cannot stand behind, so I will not.
Nothing moves forward without money. The site is a holding page now. The blog and the waitlist stay up. That is it.
I am re-evaluating over the next few days. No relaunch date, no roadmap I cannot pay for. If something changes, funding or real revenue, it gets said here first. If nothing changes, I will say that too.
Thanks to everyone who tried llmdeal. The idea was not the problem. Paying to do it properly was.