The unit economics seem tough at that price for a 1T parameter model. Even with MoE sparsity you are still VRAM bound just keeping the weights resident, which is a much higher baseline cost than serving a smaller model like Haiku.
Go to kimi chat, there will come up multiple suggestions of use cases. One of them will be the bargain robot. If you download their mobile app, the challenge to bargain will probably popup too!
Depending on how well you bargain with the robot, you can go as low as 0,99$ (difficult). Either way, their moderate plan doesn’t have to be 20$. The agent wants a good reason for why it should lower the price for you.
