Yeah even the Chinese open models have a problem that inference costs for these aren't that cheap. The only way out for the AI bubble collapse is simply more efficient hardware at lower costs and infrastructure setup downtime.

It’s just an introduction price to speed up adoption for the rest of the month, hardly worth mentioning compared to subsidized coding plans.

We know DS runs profitable, they also indicate in their paper they expect prices to drop as they get access to the next gen Huawei cards.