CodeLazr values and beliefs
As the dude who made CodeLazr (and continues to develop it), I thought I'd hand-write what some of my personal values & beliefs are, so you know what to expect from CodeLazr. That in itself is part of it: I believe in CHOICE and disclosure. More on why I built it in the first place on /story.
I believe that workflows (in general, not just AI-based ones) should be:
Open
You should be able to use the AI model that fits the job, not the one a vendor picked for you. CodeLazr works with any OpenRouter model, free ones included, and lets you bring your own key from other providers.
Honest about what works
Every test batch we run is public, including the failures and the bugs we found in our own detection logic. The limitations page says plainly what CodeLazr cannot do.
Secure
Secure in the everyday business sense, not only the network sense: your code stays in your GitHub repo, bring-your-own keys are encrypted at rest, and the AI never commits without your review unless you turn on auto-accept.
Low cost or free
A useful AI coding workflow should not require a $20+ monthly subscription. You can use CodeLazr free forever with free models or your own key, and paid plans are simply an AI budget.
Reliable
Cheap and free models fail in predictable ways. CodeLazr is built to catch those failures, retry with other models, and fix real build errors instead of leaving you to babysit every run.
Easy to use, from any device
No install, no local toolchain, no clone. If you have a browser, including on a phone or tablet, you can make a change to your repo.
See how that plays out in practice: the public test results, the honest limitations, and pricing.
Try it yourself →