Output Token Limit When Running at Cost

I am begging anyone on this planet to come up with an IDE similar to Cursor that doesn’t limit the context or output token length when using my own API Key. I understand that you all want to cut your costs because you are a company that exists to make money. That’s fair. But the fact that you severely limit the context window and the output token limit when I am using my own API Key infuriates me to no end. I pay $40 a month for this. I still have to use my own API Key, and even then, you undercut my ability to get the most out of these models. It is utterly ridiculous. I will jump ship to a different product when a viable competitor emerges. I wish PearAI were halfway decent, and then I would never have to deal with your company again. Until then, I am stuck paying money to a company I despise.