Aaron O'Leary

🛠 Has anyone built a production app with bolt.new?

I’ve been playing around with it, and it’s wild how fast it spins up a full-stack app from a prompt. But I’m wondering:

  • How well does it generate code for real-world use?

  • Any prompting tips to get better results?

  • How does it compare to others like Lovable?

Also, do you think tools like Bolt (full app generation) are the future, or will the Cursor-style coding assistant approach win out?

Would love to hear from anyone who's pushed this beyond a test project! 🚀

2.9K views

Add a comment

Replies

Best
Hussein

Haven’t tried Bolt or Lovable yet, but I’ve been using Cursor and it’s already pretty impressive. Sure, you need to know your way around code, but the speed boost is crazy. It’s wild how much AI-assisted coding has improved—definitely curious to see how full app generation tools compare! 🚀

Dineshan

Bolt is impressive but i wouldn't suggest for any complex project. Better to use ide like Cursor

Cathy Chang

Hey there! It's so cool that you've been exploring this tool and are impressed by how quickly it whips up a full - stack app from a prompt. Let's dig into your questions:

How well does it generate code for real - world use?

I've seen some pretty amazing things come out of these code - generating tools, but they're not perfect for real - world scenarios just yet. While they can create a solid foundation, the code often needs a bit of TLC. For instance, security is a biggie. In a production environment, you need to be extra cautious about things like data breaches. The generated code might not have the most robust security measures baked in right off the bat. Also, when it comes to performance optimization for high - traffic apps, it usually falls short. But don't get me wrong, it's a fantastic starting point, and with some developer know - how, you can turn that generated code into a real - world winner.

Any prompting tips to get better results?

I've found that being super - specific is key. Instead of a generic "make a social media app," go for something like "Build a React - based social media app with user profiles, a feed that displays posts in chronological order, and the ability to like and comment on posts. Use Firebase for the backend for user authentication and database storage." The more details you add, the closer the generated code will be to what you actually want. Another tip is to break down your prompts. If you're building a complex app, prompt for each major feature separately. It gives the tool a clearer picture of what you need and can lead to better - structured code.

How does it compare to others like Lovable?

This is a tough one! Each tool has its own strengths. Lovable might be great at generating code with a certain style or for specific types of projects. If it's been trained on a lot of data related to, say, mobile app development for a particular platform, it could have an edge in that area. But the tool you're using sounds like it's really fast at spinning up full - stack apps, which is a huge plus. It might be a matter of testing them side - side on the same project to see which one gives you the best results. And of course, user experience also plays a big role. If one tool has a more intuitive interface and better error handling, that can make a world of difference.

Do you think tools like Bolt (full app generation) are the future, or will the Cursor - style coding assistant approach win out?

I think we're going to see a bit of both. Tools like Bolt are a game - changer for entrepreneurs and small teams who want to quickly get an MVP out there. They can save so much time in the early stages of product development. But Cursor - style coding assistants are also invaluable for professional developers. They're like a super - smart coding buddy that can help you write better code faster, catch bugs, and suggest improvements. In the future, we might even see these two concepts merge. Imagine a full - app - generation tool that has an integrated coding assistant to help you customize and optimize the generated code. That would be a dream come true! I'm really excited to hear from people who've taken this tool beyond the test - project phase. There's so much we can learn from your experiences! 🌟

Lesley Liu

I love bolt.new and i used it to build a calendar app.
The reason i choose bolt.new over other Ai coding products is because i know nothing about coding and bolt as a full app generation tool allows me to create a product in this way.

Olive Sen

In my experience, Bolt hallucinates after the basic version is built and cannot debug properly, utilizing a lot of tokens in creating errors and fixing errors. It is superb otherwise, till now the best way to build it to use these tools as a coding assistant, than full fledged tools, unless what you need is something basic.

But the future looks promising.

Sparsh Mehta

I've tried building production apps in Bolt without success. Its good for ramping up a starting version pretty quickly, but given its so hard to debug I've stopped using it altogether. Instead I just use Composer as it does a similarly good job of getting to a starting point quickly.

James Kahwai

I have used bolt to build out really functional things like this loan calculator here; https://zingy-zabaione-42d435.netlify.app . Bolt really needs a user who's great at prompt engineering to make sure your tokens meet your intended ROI or ROR(Return on research). I have further deployed bolt.diy in my local environment and this has been the best hack. As a SE, bolt.diy does the heavy lifting and gets me an MVP(frontend with lightweight backend). I then ship the project to bolt.new where I finesse everything that wasn't complete.

I have a subscription for Lovable after going through lots of reviews which I will be building something this week then I can have a concrete comparison.