The linked page only compares to very old and very small models. But the pricing is higher even than the latest Gemini Flash 2.5 model, which performs far better than anything they compare to.
If speed is your most important metric, I could still see there being a niche for this.
From a pure VC perspective though, I wonder if they'd be better off Open Sourcing their model to get faster innovation + centralization like Llama has done. (Or Mistral with keeping some models private, some public.)
Use it as marketing, get your name out there, and have people use your API when they realize they don't want to deal with scaling AI compute themselves lol
They're comparing against the fastest models. That's why smaller models are shown.
Put another way, how much would company x be willing to spend on "here's a repo, here are the tests, here is the speed now, make this faster while still passing all the tests". If it "solves" something in cudnn that makes it 10% faster, how much would nvidia pay for this? 1m$? 10m$?