I'll keep things brief, there are two things I hate about using AI coding models: paying a subscription and how bad they are for the environment.
So my cofounder and I built Bound. Simply put, we provide Cursor's "tab" feature but instead of using a cloud model we use an ultra-small local LLM trained on your own repo (that and a couple other tricks to make sure it's the same quality as cursor). The nice thing here is that you own the model and it's so small that it runs on your laptop with no latency issues.
Join the waitlist on our website and we'll get in touch to implement it for you.