Replit drove me mad!
I started using Replit about 3 weeks ago. First on a trial basis and then decided to pay the quite ok monthly fee of $25. What I didn't realize back then was that once they got my credit card info they just keep charging for use even if cancel my sub.
The thing is that Replit charges 'checkpoints' a 25 cents each no matter what it does. Implementing Google Auth with 2FA or failing over and over again to format an output i Typescript, same result. The checkpoints just runs away. When the AI fails to solve a very basic variable naming problem for the Xth time you tend to get a bit frustrated - to say the least. Tell the AI to mind the difference between CamelCase and snake_case and it remembers it for less than 5 minutes. Putting it in writing in the manifest Instructions.md makes no difference.
After about a week I decided to make a first deployment. The AI ran all sorts of test and then deployed. Successfully in its own words. Except not a single endpoint worked in deployment. Not even logging in to the Express backend as admin. So I asked the AI if it thought it was ok that code that worked in the Replit dev env did not work in their own deployment env (based on Google Cloud). It wasn't. So I sent a mail to support. Which yielded absolutely nothing but a boilerplats BS answer. You see, the support too is an AI. How clever is that!
Well I toiled on and after about two weeks and some $250 worth of checkpoints I had a working web for matching local contractors with local property owners. A hyper local matching service. Then when I'm running the last tests before my third deployment I suddenly notice that the admin password is changed back to a very unsafe password I used the first days of developing. Looking through the pages of my app (not too many in total) I realised that the AI had reverted back to several days old code. When I asked to have the code from last night back the AI happily restored a 2 weeks old backup taking me even further back. It then told me that it had not changed anything in the code, that someone else must have changed the code and that there were no newer code available.
This is when I gave upp and told the AI to go F*CK itself.
So unless Replit gets back to me and provides my latest code back ... or credit me at least $200 I will have to write this company off as complete fraud. DO NOT PAY for the Replit service. It's taken the company 9 years (founded 2016) and 221 million dollars worth of venture capital to get to this state of uselessness. Flabbergasting!
Replies
Hey, just saw this and sorry about the inconveniences you've had using Replit.
Happy to help you recover your code and get you back to a working state. Please shoot me an email kody.low@replit.com and I'll schedule a call with you.
Product Hunt
That sounds really frustrating. My experience with Replit and most recently @Windsurf has sometimes been magical and sometimes feels like banging my head against a wall. It's especially frustrating when the AI seems to create more problems instead of fixing them and also charges you more in the process.
I don't think these companies are frauds though. I just think that AI coding tools are really hard to build and currently really hard to use. I think it's really easy to make a demo app that does about 80% of what you'd want... but it's really hard to get that last 20%.
Kody from the Replit team responded below - he might be able to help.
Also check out some of these other forum threads that have been helping me personally get better at vibecoding:
#vibecoding: What are your favorite Cursor pro-tips?
Vibe coding tips & tricks and "Vibe coding" for non-coders both by @dan_leshem
What's your Replit "vibe coding" workflow? (I'm struggling here...) - a thread I started to get help, and @kody_low responded with some excellent advice
@rajiv_ayyangar Great references
Random Picker
Wow. That sounds frustrating. My experience has been the complete opposite and I consider Replit indispensable.
I've been using it for over 2 years. Without AI, it's already an excellent browser-based IDE. With AI ... it's been an amazing tool for me. Transformative.
Being code-inclined, I rarely use the Agent and instead use the assistant ... I find it works amazing when collaborating with AI within the code. Plus, assistant advanced iterations are $0.05 and basic ones are free. (code vs. text or copy changes). Better results, cheaper, comfortable fit in my workflow.
I agree with the other comment that the AI-Builder problems (no-code) are likely not unique to Replit.
Last thing I'll say, the customer service has been exceptional when I have had issues that required explanation, especially around usage charges. I'd be surprised, based on my experience, if they couldn't make things right for you.
Curious ... if you have it handy, what was your first initial prompt?
For what it's worth, a lot of what you are describing isn't necessarily a Replit problem but a problem with most of the current generation of these AI builder tools, including Lovable and others.
It's definitely frustrating, and I have had the same experience with the AI, strangely going back several revisions or, as you say, days, and undoing 5 things we already fixed while fixing the 1 thing we are currently working on.
I also find across AI that it will often ignore the hard-written instructions or manifest file and go off on a walkabout.
I can't comment on your payment struggles beyond saying there's no excuse for a company to continue billing you if you have cancelled.
It's definitely sucks when these companies are growing their user bases so fast that they don't care about the churn of losing earlier adopters by offering poor support.
Good work on your app, though. Hopefully you can get it working 100%.
I'm definitely not getting my startup replit credits now 😅
Mailmodo
Have faced something where the Agent keeps on doing the same thing over and over again like if you ask it to check console logs and fix issues. Many times, it is neither able to detect well nor ask the user for any insights that might help it. I have spent many dollars fixing an issue with the AI agent/assistant, which only got fixed when I peered into the code myself.
Do you know if lovable or bolt are better at this?