Firebase Studio(firebase.studio)
firebase.studio
Firebase Studio
https://firebase.studio
118 comments
I built a complete working application (errortexts.com) using an AI tool, so I have a little insight on this.
At first, the product I was using (lovable.dev) seemed to me exactly as you described. I gave it a basic app outline and hit run, and it produced something that superficially looked right but did nothing.
So I asked some other people for advice, and they said you have to hold its hand and go step by step. So I did.
I told it, give me a landing page that matches [product description], but implement nothing else. Then, ok, let's set up auth - add a sign in and sign up dialog. Then, ok, let's create a user account page. Bit by bit.
It succeeded wildly. I was able to build the whole thing in 3 days. I'm not capable of that on my own, it would have taken me 3 weeks. Sometimes the AI got stuck and I had to manually go in and accomplish what I wanted. It took over 100 steps to complete the product, and probably around 10-20 times I had to revert its changes and give it more specific instructions. I had to check its work at every iteration, just like with a junior developer.
But it worked. And it's going to get better. Would I use this for "something important"? Depends how you define that. I used it to build a working product. Would I start letting it modify an existing mature codebase willy-nilly? No, probably not. Would I let it write cryptographic logic or trust that it wrote bulletproof code from a security standpoint in a sensitive context? No.
But for a simple application, it was an incredibly powerful tool. Especially for something that didn't even exist just 2 years ago. Give this a decade and it's going to change all our careers even more than it already has.
At first, the product I was using (lovable.dev) seemed to me exactly as you described. I gave it a basic app outline and hit run, and it produced something that superficially looked right but did nothing.
So I asked some other people for advice, and they said you have to hold its hand and go step by step. So I did.
I told it, give me a landing page that matches [product description], but implement nothing else. Then, ok, let's set up auth - add a sign in and sign up dialog. Then, ok, let's create a user account page. Bit by bit.
It succeeded wildly. I was able to build the whole thing in 3 days. I'm not capable of that on my own, it would have taken me 3 weeks. Sometimes the AI got stuck and I had to manually go in and accomplish what I wanted. It took over 100 steps to complete the product, and probably around 10-20 times I had to revert its changes and give it more specific instructions. I had to check its work at every iteration, just like with a junior developer.
But it worked. And it's going to get better. Would I use this for "something important"? Depends how you define that. I used it to build a working product. Would I start letting it modify an existing mature codebase willy-nilly? No, probably not. Would I let it write cryptographic logic or trust that it wrote bulletproof code from a security standpoint in a sensitive context? No.
But for a simple application, it was an incredibly powerful tool. Especially for something that didn't even exist just 2 years ago. Give this a decade and it's going to change all our careers even more than it already has.
Can you... provide the prompts? Because I have a hard time believing this.
I have tried the hand holding approach with Cursor. It doesn't work for me. I have to constantly correct and over correct. Getting auth working sounds insane to me.
I have tried the hand holding approach with Cursor. It doesn't work for me. I have to constantly correct and over correct. Getting auth working sounds insane to me.
What exactly surprises you? You should try Gemini 2.5 Pro in Google AI Studio. Set the temperature to around 0.3, and in the system prompt, tell it to only edit exactly what you ask for and nothing else.
This model works really well. For example, simple things like [1], [2], and [3] can apparently be generated with just a couple of prompts.
[1] https://koreanrandom.com/en/games/through-the-space/
[2] https://koreanrandom.com/en/games/tanchiki/
[3] https://koreanrandom.com/en/games/sharik/
According to the author, these were made with Gemini 2.5 Pro without any manual coding by a human.
Cursor isn't as powerful as Gemini in AI Studio because AI Studio gives you full control over the model's settings and how it processes code.
Plus, the massive 1 million token context window is incredibly helpful for working with large codebases. You can use tools like code2prompt and repomix to feed all the necessary context into AI Studio from the clipboard for those projects.
This model works really well. For example, simple things like [1], [2], and [3] can apparently be generated with just a couple of prompts.
[1] https://koreanrandom.com/en/games/through-the-space/
[2] https://koreanrandom.com/en/games/tanchiki/
[3] https://koreanrandom.com/en/games/sharik/
According to the author, these were made with Gemini 2.5 Pro without any manual coding by a human.
Cursor isn't as powerful as Gemini in AI Studio because AI Studio gives you full control over the model's settings and how it processes code.
Plus, the massive 1 million token context window is incredibly helpful for working with large codebases. You can use tools like code2prompt and repomix to feed all the necessary context into AI Studio from the clipboard for those projects.
I just got auth working with JWT through total vibecoding, using Claude+RooCode. Other bits of the app I needed a couple of tries, but auth worked immediately. I guess these models have seen express + node + JWT a million times.
I don't have access to the precise prompt, but I told it something like "only implement basic authentication based on JWT, using just email + password.", then asked it to add a simple registration form, then the password reset flow... step by step, but with little guidance.
At every prompt I review the changes on git, and commit.
I don't have access to the precise prompt, but I told it something like "only implement basic authentication based on JWT, using just email + password.", then asked it to add a simple registration form, then the password reset flow... step by step, but with little guidance.
At every prompt I review the changes on git, and commit.
I have had Claude Code build authentication before successfully, following a pattern approximately like this: https://harper.blog/2025/02/16/my-llm-codegen-workflow-atm/
I love the Java code sample:
```
// It works, but we have no idea how to send an HTTP request in java.
// If you do, send us a nice example to [email protected].
// You just need to POST JSON with the keys `api_key` and optionally `message`.
```
That's brilliant. But it does make me wonder why the LLM couldn't have provided you a suitable Java code example?
```
// It works, but we have no idea how to send an HTTP request in java.
// If you do, send us a nice example to [email protected].
// You just need to POST JSON with the keys `api_key` and optionally `message`.
```
That's brilliant. But it does make me wonder why the LLM couldn't have provided you a suitable Java code example?
I wondered the same. Those code examples took an hour or so to figure out. It's a very basic exercise, sending an HTTP request, but doing it in nine languages, the majority of which you've never used, requires a little research. The llm can generate them all, but I had to run them locally to make sure they worked, and they oftentimes didn't. Most of them were easy and only took a minute or two to figure out, but after an hour and a half I gave up on Java. I literally could not figure it out. I'm sure I could have if I went long enough, but at some point you just have to cut your losses and decide it's not worth your time. This was not going to make or break my product offering, I'm not trying to learn Java, and I just wasn't interested in spending more time trying to figure it out. The llm was happy to generate examples, but they didn't work.
With these tools you have to talk to them as if you were talking to a knowledgeable, but a bit clueless junior developer. Sometimes it's almost as if you were coding it, just without typing the code.
I put your prompt into the vibe coding tool I'm working on (shameless plug).
The first version[0] looked good, but when I inspected it I found that it just picked an I Ching prediction at random on the back-end, instead of actually flipping coins.
I updated the prompt to the following:
> Create app where you input a question, then flip coins to generate an I Ching prediction (client-side). First show the standard I Ching prediction and it's hexagram, and then use AI to generate a fortune analysis based on the prediction and your initial question.
And the result was much more laborious[1] of a UI :shrug:
[0]https://iching.magicloops.app
[1]https://iching2.magicloops.app
The first version[0] looked good, but when I inspected it I found that it just picked an I Ching prediction at random on the back-end, instead of actually flipping coins.
I updated the prompt to the following:
> Create app where you input a question, then flip coins to generate an I Ching prediction (client-side). First show the standard I Ching prediction and it's hexagram, and then use AI to generate a fortune analysis based on the prediction and your initial question.
And the result was much more laborious[1] of a UI :shrug:
[0]https://iching.magicloops.app
[1]https://iching2.magicloops.app
The ai part of the app is basically useless. After 2 hours of “vibe coding” a chess clock flutter app I got basically nothing in the end.
It broke more and more each message. I tried fixing stuff myself but it would mess it up again. Would not recommend anyone to use it.
Now for the non ai part: super cool. I love the nix environment. Its fascinating how they handle the previews for example. I got geekbench up and running an the cpu is a bit worse than an iphone 15 pro max, but it has 32 gigs of ram!
It broke more and more each message. I tried fixing stuff myself but it would mess it up again. Would not recommend anyone to use it.
Now for the non ai part: super cool. I love the nix environment. Its fascinating how they handle the previews for example. I got geekbench up and running an the cpu is a bit worse than an iphone 15 pro max, but it has 32 gigs of ram!
[ I work on Dart at Google. ]
The app prototyping logic in Firebase Studio isn't wired up for Flutter/Dart yet. You can play with Gemini+Dart/Flutter here: https://dartpad.dev/?channel=main.
We're working with the Firebase Studio team to integrate. FWIW, it seems to do fairly well with "Create a chess clock app".
The app prototyping logic in Firebase Studio isn't wired up for Flutter/Dart yet. You can play with Gemini+Dart/Flutter here: https://dartpad.dev/?channel=main.
We're working with the Firebase Studio team to integrate. FWIW, it seems to do fairly well with "Create a chess clock app".
Hey. That makes more sense. To be honest, I don’t think it should be one of the main things on the front page when creating a new project in that case.
However for what it’s worth Ive also asked gemini to translate this simple chess clock app to react because of how buggy the flutter project was and it just got stuck again and again no matter how much I helped it
[Full disclosure, work @ Firebase]
But try this:
Open up a blank Flutter template in Studio: https://studio.firebase.google.com/new/flutter
After everything initializes and your Android and web previews set up, open chat by clicking the little Gemini spark at the bottom of the workspace and then add your prompt.
YMMV, but I got a very basic, but working chess clock in one shot with "Can you replace this sample Flutter project with a fully-functional chess clock that works on Android and on the web?"
But try this:
Open up a blank Flutter template in Studio: https://studio.firebase.google.com/new/flutter
After everything initializes and your Android and web previews set up, open chat by clicking the little Gemini spark at the bottom of the workspace and then add your prompt.
YMMV, but I got a very basic, but working chess clock in one shot with "Can you replace this sample Flutter project with a fully-functional chess clock that works on Android and on the web?"
That's how I've done it when it did not work. I did, however, go easier, one by one. I first made a simple countdown. Then I made it count down from 10 minutes to 0. Then I created 2 of them. Then I put them side by side. Then I did the logic, one counterstarting each other. Every step of the way, it was filled with loads of errors and loops out of which the ai couldnt get out, to the point where I would need to go there and fix it myself. Errors ranged from logical, to super simple syntax issues, such as missing colons, or brackets.
I have tried again step by step with what you said here, copy pasted the prompt even and after a couple changes here and there it loses the plot and gets stuck and seemingly no amount of prompting can get it out.
To make sure I tried 3 times in a row. Same result.
I feel like in the emperors new clothes story.
To make sure I tried 3 times in a row. Same result.
I feel like in the emperors new clothes story.
Well, not conpletely Emperor’s New Clothes: after I got my basic chess clock, I asked for skeuomorphic changes…I got a fantastic “next steps” summary in return…but when I said “Go on ahead, your choice,” it created an incompatible dependency nest trying to add clicky sounds.
Gemini actually dug me out of it, though, with additional prompting (this kind of surprised me; I assumed I’d have to eventually really step in). After all that distraction, though, I had to reset the convo to get it to focus on the files again.
Totally understand that you might be too annoyed at this point, but if you do have a few minutes, it’d be amazing to file a case @ https://firebase.google.com/support/troubleshooter/studio for the team to dig into.
Gemini actually dug me out of it, though, with additional prompting (this kind of surprised me; I assumed I’d have to eventually really step in). After all that distraction, though, I had to reset the convo to get it to focus on the files again.
Totally understand that you might be too annoyed at this point, but if you do have a few minutes, it’d be amazing to file a case @ https://firebase.google.com/support/troubleshooter/studio for the team to dig into.
As a person currently working on a project that uses Firestore (the db component of Firebase), there is one thing - and only one thing - I want.
A web GUI for Firestore that lets me work on documents like, idk, any other DBMS GUI would: the ability to select multiple records, and operate on them.
That's literally it. I don't need AI, I don't need dark mode, I don't even need MongoDB compatibility. I just want to select multiple documents with my mouse and do things to them.
A web GUI for Firestore that lets me work on documents like, idk, any other DBMS GUI would: the ability to select multiple records, and operate on them.
That's literally it. I don't need AI, I don't need dark mode, I don't even need MongoDB compatibility. I just want to select multiple documents with my mouse and do things to them.
I built exactly this in less than a day using Windsurf.
Come up with your data model, explain it to the AI, and tell it to give you a CRUD for it.
I'm pretty sure it's perfectly doable to ask it to give you a dynamic crud based on the "shape" of the data in Firestore.
Sadly it's an internal tool for work, so I can't share.
Come up with your data model, explain it to the AI, and tell it to give you a CRUD for it.
I'm pretty sure it's perfectly doable to ask it to give you a dynamic crud based on the "shape" of the data in Firestore.
Sadly it's an internal tool for work, so I can't share.
For me the biggest missing block is the text search API. It's ridiculous that you can't add a basic search input to your Firebase-based website unless you use TypeSense, Algolia or some other additional database that you have to manage and keep in sync.
Despite all the recent enshitification, I can't think of any alternative solution that would come even close to what Firebase has to offer. Authentication API is especially hard to beat (cheap and very easy to integrate).
Despite all the recent enshitification, I can't think of any alternative solution that would come even close to what Firebase has to offer. Authentication API is especially hard to beat (cheap and very easy to integrate).
Check out fuegoapp.dev — it comes with a bunch of handy tools to manage Firestore and Firebase users
Have you tried Rowy.io?
I've generally looked around.
Looking at that, I'm not sure corporate would be down for something like that as a solution for "select multiple documents in a database GUI and operate upon them". Could be wrong though.
Looking at that, I'm not sure corporate would be down for something like that as a solution for "select multiple documents in a database GUI and operate upon them". Could be wrong though.
I’ve enjoyed firefoo for this
It's a rebranding of https://idx.google.com/
I was under the impression that idx was formerly just the editor, but I guess that's changed.
First off, this looks really cool and I'm excited to see more things like this.
The overall chat in the HN conversation has got me thinking, though.
Around 7 years ago in my career, one of my most common actions for one-off scripts was for me to create a WinForms application with, often, a couple text boxes and a "Run" button of some sort.
The text boxes would be the inputs and the run button would ... run. There was also often like a text output or bunch of loglines or something. I wrote almost exclusively in C# at the time, so it was a way to shove a bunch of C# code into place and test it.
I did this for random and arbitrary things I needed to process or solve, a lot like how I used Python or Ruby in the future.
I bet it's actually pretty common for people to need "a script that does a thing", and I think, maybe, that's where a lot of the AI scripting of the most immediate use is going to be. If it can be a familiar interface for people to build (in the past, the IDE) and a familiar or simple place to interact with the generated script (the WinForms + buttons), these programs to generate scripts and do "stuff" could likely spread pretty wide.
I think Jupyter Notebooks are another example of this, another precursor, of sorts?
The overall chat in the HN conversation has got me thinking, though.
Around 7 years ago in my career, one of my most common actions for one-off scripts was for me to create a WinForms application with, often, a couple text boxes and a "Run" button of some sort.
The text boxes would be the inputs and the run button would ... run. There was also often like a text output or bunch of loglines or something. I wrote almost exclusively in C# at the time, so it was a way to shove a bunch of C# code into place and test it.
I did this for random and arbitrary things I needed to process or solve, a lot like how I used Python or Ruby in the future.
I bet it's actually pretty common for people to need "a script that does a thing", and I think, maybe, that's where a lot of the AI scripting of the most immediate use is going to be. If it can be a familiar interface for people to build (in the past, the IDE) and a familiar or simple place to interact with the generated script (the WinForms + buttons), these programs to generate scripts and do "stuff" could likely spread pretty wide.
I think Jupyter Notebooks are another example of this, another precursor, of sorts?
That's one use case where LLMs really help me, one-off scripts where I know exactly what I need done, know it's possible but would take me much longer to brush off my Bash, Python, etc. skills to write it. Give the LLM a prompt, let it write the scaffolding, do the tweaks I need, and iterate over with the LLM if I forgot how exactly it was to write a for-loop in Bash.
Software engineers, who are the most skilled in terms of holding ai's hand to create a product, should be cloning every single saas out there and making money by eating a share of the market. AI is a great way for engineers to become founders. Let's bring the competition.
5 years working at startups has taught me that go-to-market is almost always the most important function. There are very few truly novel spaces and apps. You can clone whatever you want, but you'll get nowhere without a solid GTM.
That applies to new products not clones. AI can also help with that thanks.
This especially applies to clones. How do you stand out and find the right customer niche?
Eating the big fish is one great book about competing with a market leader. You're talking like go to market is rocket science. Anyone can read a book and follow the steps. Many don't.
> You're talking like go to market is rocket science.
That's all good. I didn't say cloning a saas using AI is everything that's needed. Once a software engineer has cloned a saas they can do go to market science, if that' what you want to call it
If you think that latter part is easy, then you should definitely go clone a SaaS and exit with millions. You have all the tools at your disposal to clone any app that you want now with minimal effort. Be a 1 man team, no overhead. Undercut on price and exit.
In fact, it's not that easy. The second part -- GTM -- is the hardest part; even if you could clone Notion, getting that product in front of customers is the hard part.
In fact, it's not that easy. The second part -- GTM -- is the hardest part; even if you could clone Notion, getting that product in front of customers is the hard part.
I'm sorry you had a tough time GTM'ing
Hey, if you're a GTM genius, let's do it! Over the last 5 years, I've worked at an $8m seed startup that couldn't get it right and folded last Dec. I've built three other products that didn't go anywhere because of GTM (lack thereof). If you're the genius I've been looking for, then ping my LI!
Let's go brrrr! I always wondered how the end of yc would look like.
YC is money laundering at this point. "Unicorns" and "disrupting" are just other names for money laundering and illegal activity.
I was excited to try this out because I've had a lot of trouble getting the Supabase integrations to work on Lovable and Bolt.new.
Sorry to say that Firebase Studio did an awful job. It did not successfully build even the first view of the app I asked for. It feels like I'm stepping back to release day of GPT-4.
Am I missing a switch to use the good Gemini 2.5 somewhere? I could tell from their response speed that I was not using a thinking model.
Sorry to say that Firebase Studio did an awful job. It did not successfully build even the first view of the app I asked for. It feels like I'm stepping back to release day of GPT-4.
Am I missing a switch to use the good Gemini 2.5 somewhere? I could tell from their response speed that I was not using a thinking model.
this isn't new.. (edit, well actually they've added the "prompt from scratch" thingy like v0.dev/replit/lovable that idx.dev didn't have before) wasn't really gaining the traction they wanted so decided to rebrand. not a bad idea. I use idx.dev (now firebase.studio) and like it. They may not have a "template" for the exact stack/app type you are working with, but it's pretty easy to just setup a blank workspace and modify the nix.dev file as needed..
Here's a dev.nix file that works for FastHTML, which is a little tricky to get the preview working but using this dev.nix it works out of the box. Would probably work with any Starlette or Uvicorn based app as well (obviously make sure you have requirements.txt in order).
https://community.firebasestudio.dev/t/working-fasthtml-a-ne...
https://community.firebasestudio.dev/t/working-fasthtml-a-ne...
How was your experience? Would it compare to Cursor, Windsurf, Cline/Roo?
Friend is working on a SaaS startup solo as a non-technical founder. Been building using Replit and Cursor with Claude 3.5 and 3.7 and launched with live customers.
I preface because he's been using AI basically every day to build his SaaS.
His comments today:
I preface because he's been using AI basically every day to build his SaaS.
His comments today:
> also gemini 2.5 is beating claudeI've been enjoying just trying to one-shot little pages that I think are useful with v0.dev. It does a pretty good job and I like that I can just one-button deploy to my Vercel account (which "just works"). Here's one I did today:
https://v0-github-repo-stream.vercel.app/
https://v0-github-repo-stream.vercel.app/
One more for fun :)
https://retro-computing-stream.onrender.com/
https://retro-computing-stream.onrender.com/
Mind revealing the product/site or are they going to do an upcoming hn launch..?
He has asked to not reveal it and not likely to reveal it. I asked him about it because I think the press would be insane, but it's a niche B2B SaaS and he doesn't want the "made with AI" branding to hurt the image of the product from a customer perspective. Second concern is people trying to hack it in case the LLM took shortcuts on security.
But I can say that as a 20yr+ eng. and having worked with him on a prior startup, I'm super, super impressed how far he got by himself in under 4 months. Basically a full-blown industry specific CRM with Stripe integration, OCR from mobile web app, desktop web app, marketplace. He himself is concerned that it's the end of SaaS as he thinks anyone can build these things now, but I think there's still a bit of a chasm when it comes to some ancillary pieces like databases.
But I can say that as a 20yr+ eng. and having worked with him on a prior startup, I'm super, super impressed how far he got by himself in under 4 months. Basically a full-blown industry specific CRM with Stripe integration, OCR from mobile web app, desktop web app, marketplace. He himself is concerned that it's the end of SaaS as he thinks anyone can build these things now, but I think there's still a bit of a chasm when it comes to some ancillary pieces like databases.
> he doesn't want the "made with AI" branding to hurt the image of the product from a customer perspective
That's kinda telling.
That's kinda telling.
I ran into a Show HN thread not so long ago where people were outright hostile to a minimally technical developer that did something actually impressive, in my eyes at least, but he was open about using AI for everything. I've not seen a more badly mannered thread on HN.
It's like some people are triggered by non-veterans programming with AI. I think everyone understands the downsides. Why get angry about it?
It's like some people are triggered by non-veterans programming with AI. I think everyone understands the downsides. Why get angry about it?
Yeah. It seems t me though that there two distinct demographics at play here. The"HN commenting crowd" and the "potential customers for a SaaS".
I understand and (mostly) stay out of the arguments between HN users (both "technical" and "non-technical" business types) about AI coding. (And I'm not particularly interested in airing those grievances again in this subthread, well mannered or not.)
But I'm intrigued that the dev mentioned upthread is intentionally hiding his use of LLM coding, instead of joining in with the AI buzzword bandwagon jumping that's going on pretty much everywhere I look in tech, from tiny to behemoth corporations.
I understand and (mostly) stay out of the arguments between HN users (both "technical" and "non-technical" business types) about AI coding. (And I'm not particularly interested in airing those grievances again in this subthread, well mannered or not.)
But I'm intrigued that the dev mentioned upthread is intentionally hiding his use of LLM coding, instead of joining in with the AI buzzword bandwagon jumping that's going on pretty much everywhere I look in tech, from tiny to behemoth corporations.
It is inconceivable that someone would become angry on the internet. You're going to turn off the computer and run outside as fast as possible if you sense that your emotions are even just ever so slightly shifting from enjoyment.
Sarcasm is the internet default, though. Perhaps you didn't quite pick up on it that time?
Sarcasm is the internet default, though. Perhaps you didn't quite pick up on it that time?
If promoting your app as AI-made is going to draw ire regardless of whether it makes sense to, then it's just pragmatic business not to do so. You gain nothing by it. Doesn't have anything to do with being upset or offended by internet comments. Not all criticism is sensible.
The question asked why would people get angry about AI on an internet forum? The answer is that they don't.
What does that have to do with promoting apps and whether or not it is sensible business decision to do so? You must have accidentally hit the wrong reply button? I think I saw that discussion taking place elsewhere.
What does that have to do with promoting apps and whether or not it is sensible business decision to do so? You must have accidentally hit the wrong reply button? I think I saw that discussion taking place elsewhere.
>The question asked why would people get angry about AI on an internet forum? The answer is that they don't.
But they do lol.
But they do lol.
No they don't. Computers have a power button. If there was even the slightest indication that you were no longer having the time of your life, you'd press it and find something more enjoyable to do.
But as people on the internet do find great enjoyment in being funny (for some definition of funny), sometimes their humour might come across as being "angry" if you don't understand the context in which it is taking place.
But as people on the internet do find great enjoyment in being funny (for some definition of funny), sometimes their humour might come across as being "angry" if you don't understand the context in which it is taking place.
Sarcasm or cynicism is the default for some people and some communities. That's fine if it works for them. I'm not interested in it, and the HN I know has little of it. When I said "hostile" I wasn't talking about either of these though.
At the same time, he's really helping small business customers by offering a competitive product at a fraction of the price of the dominant vendors in the space (a duopoly) because he has no overhead. Customers love the product. No ads or marketing and he's got 30ish customers via word of mouth.
Interesting but a CRM built by a non-technical person using AI?
I think that's pretty irresponsible - building a product that holds PII without any concern for security. Not to mention that's a lawsuit waiting to happen.
I think that's pretty irresponsible - building a product that holds PII without any concern for security. Not to mention that's a lawsuit waiting to happen.
What stack is he using for all of it? I’m curious which tools or languages AI tends to be the most effective with?
Tooling: started developing in Replit including hosting, added Cursor later when it got a bit more complicated.
FE: React SPA with Tailwind; I don't think he was knowledgeable enough to pick a UI component library.
BE: Database is Supabase, auth is Firebase. He had a hard time getting the AI to work with Supabase auth so I suggested Firebase instead (Firebase auth is dead simple and lots of docs). This created another challenge which was using the claims from Firebase in Supabase for the row level security. Rest of the backend is running in Replit autoscale deployments. Stripe integration for payments and other integrations for common third party systems in this particular industry. All done via AI and handing it relevant documentation and code samples.
Replit's tooling is interesting because it can run the app and see the errors in the console. It can also take its own screenshots to see if it's getting it right. I watched some of his sessions and the LLM often makes mistakes. He'll try a few passes to get the LLM to fix the mistake, but if it doesn't work, he moves and will come back to it if it is non-breaking.
I talked to him a bit more and he said he thinks his "super power" is managing an offshore team in India for a few years so he's very used to managing features and code quality via written instructions. Where an engineer (like me) might just give up and take it into our own hands when things go wrong, he can just keep chatting with the AI and pointing it to docs and such.
FE: React SPA with Tailwind; I don't think he was knowledgeable enough to pick a UI component library.
BE: Database is Supabase, auth is Firebase. He had a hard time getting the AI to work with Supabase auth so I suggested Firebase instead (Firebase auth is dead simple and lots of docs). This created another challenge which was using the claims from Firebase in Supabase for the row level security. Rest of the backend is running in Replit autoscale deployments. Stripe integration for payments and other integrations for common third party systems in this particular industry. All done via AI and handing it relevant documentation and code samples.
Replit's tooling is interesting because it can run the app and see the errors in the console. It can also take its own screenshots to see if it's getting it right. I watched some of his sessions and the LLM often makes mistakes. He'll try a few passes to get the LLM to fix the mistake, but if it doesn't work, he moves and will come back to it if it is non-breaking.
I talked to him a bit more and he said he thinks his "super power" is managing an offshore team in India for a few years so he's very used to managing features and code quality via written instructions. Where an engineer (like me) might just give up and take it into our own hands when things go wrong, he can just keep chatting with the AI and pointing it to docs and such.
> Second concern is people trying to hack it in case the LLM took shortcuts on security.
This… warrants some self reflection.
This… warrants some self reflection.
The obvious solution is to use AI to find security holes in the AI generated app /s
tbh i haven't used the agentic/gemini integration hardly at all yet. All i did was add an already completed project just to see how the web ide/preview performed.
I tried this on Mobile Safari and it clearly wasn't designed with mobile in mind, then when it got to the preview your app screen the preview panel was blank (it worked when I switched to Firefox on macOS).
I do most of my vibe coding on my phone, so that's pretty disappointing!
I do most of my vibe coding on my phone, so that's pretty disappointing!
> I do most of my vibe coding on my phone, so that's pretty disappointing
Curious what kind of phone apps you’ve been using to do vibe coding, I suppose it auto deploys stuff also?
Curious what kind of phone apps you’ve been using to do vibe coding, I suppose it auto deploys stuff also?
Exclusively HTML+JavaScript web apps. Many of the tools listed on https://tools.simonwillison.net/colophon were built on my phone - usually directly in the Claude app, then I copy and paste the resulting HTML into the GitHub web editor on https://github.com/simonw/tools/new/main to create the page.
That repo automatically deploys to my https://tools.simonwillison.net/ site.
That repo automatically deploys to my https://tools.simonwillison.net/ site.
Your output is prodigious! Absolutely blown away by how much you've been able to write over the years
There are so many Google products that overlap it's mind boggling
They're trying to get in on the AI IDE / AI coding land rush so they don't have to acquire one of the leaders later. Or get blocked from acquiring one for antitrust reasons. Winning a chunk of this space feels like a major chess move that enables a lot of market optionality.
Google and Microsoft both have web IDEs, so it's a natural product evolution for these (though Microsoft's Github is much better positioned with its deep distribution and integrations).
It still feels like these are so far behind Cursor and Zed and the like in terms of getting real work done, but maybe the customer is new or non-engineers.
I haven't looked at startups like Replit. Are they even in this game?
And then there are the end-to-end products like Vercel's V0, Lovable, Bolt, etc. If they can get past the 80% problem, they might be a huge deal.
This is going to be quite the battle.
Google and Microsoft both have web IDEs, so it's a natural product evolution for these (though Microsoft's Github is much better positioned with its deep distribution and integrations).
It still feels like these are so far behind Cursor and Zed and the like in terms of getting real work done, but maybe the customer is new or non-engineers.
I haven't looked at startups like Replit. Are they even in this game?
And then there are the end-to-end products like Vercel's V0, Lovable, Bolt, etc. If they can get past the 80% problem, they might be a huge deal.
This is going to be quite the battle.
Do you think Cursor's AI tools are significantly better than VSCode's Copilot? I haven't had time to check out Cursor. But from the general description, the AI tooling sounds very similar.
It is a common opinion (shared by myself and many others, including the streamer Primeagen, for example) that Cursor is much better than Copilot
Google also has quite the graveyard of products. I dont trust them. Hell they even have an OS that competes directly with Android.
It's also literally named a product that has historical meaning in the developer community.
You mean Tamplin didn't ship this?! Srsly tho, I think the idea (or at least was my impression) that it's backed by firebase and firebase has just become ubiquitous for a part of the gcp backend stack? so I thought firebase studio was a way to put a v0 app creator tool on top of firebase? In that context it made sense to me, but maybe I'm misunderstanding.
And which one this overlaps?
To be honest, I don't know what this overlaps with before investing the time using it. As far as I know they already have Google AI Studio, Vertex AI Studio, Project IDX (which I suppose this replaces), Firebase Studio and each individual tool also has their own AI features/plugins (Android Studio w/ Gemini, etc)
This is renamed Project IDX. And that is something like GitHub Codespaces. Only instead of VS Code it uses a laundered version of VS Code called theia running in browser or Android studio remote video streamed as frontend and backed by a nix configured docker container.
I’m afraid it will use my google cloud billing and start charging. My prompt “replicate Salesforce”
It is free. though it takes your data for future training.
https://firebase.google.com/docs/gemini-in-firebase#:~:text=...
https://firebase.google.com/docs/gemini-in-firebase#:~:text=...
The final result is very poor. Google looks like Hooli with chaotic management. The design has lost its aesthetics, and the gradient color is very ugly.
However, the aistudio updated today is good.
So it seems that the output is so different because the person in charge is different?
From this perspective, there is no hope for Firebase, because the person in charge has poor management logic. even with the best resources in the world, it doesn't help
However, the aistudio updated today is good.
So it seems that the output is so different because the person in charge is different?
From this perspective, there is no hope for Firebase, because the person in charge has poor management logic. even with the best resources in the world, it doesn't help
I know it'll probably get better over time but it's pretty awful compared to what you can do with v0 right now.
About the editor, do they allow Flutter development in a Monaco stack? Last time I checked, it was tied to their Intellij-based IDE which needs regular force updates.
About the executables, are backend apps running on a temporary VM which can run Postgres etc? Does the Android emulator launch from web?
About the executables, are backend apps running on a temporary VM which can run Postgres etc? Does the Android emulator launch from web?
I don't know what Monaco is, but most Flutter developers these days are using VS Code as far as I know.
Firebase Studio is an agentic, cloud-based development environment that gives you powerful tools and AI agents directly in your browser. With Firebase Studio, you can prototype, build, test, publish, and iterate on full-stack AI apps from a single place.
I admit, this comment did a lot better at Buzzword Bingo compared to my attempt in ChatGPT: ("Give me a two sentence marketing pitch for Firebase Studio")
"Firebase Studio is a powerful visual development platform that lets you build, customize, and deploy full-featured apps with the speed of low-code and the flexibility of Firebase. Whether you're prototyping or scaling to millions of users, Firebase Studio streamlines your workflow with real-time data, seamless backend integration, and built-in hosting."
"Firebase Studio is a powerful visual development platform that lets you build, customize, and deploy full-featured apps with the speed of low-code and the flexibility of Firebase. Whether you're prototyping or scaling to millions of users, Firebase Studio streamlines your workflow with real-time data, seamless backend integration, and built-in hosting."
Oof, “low-code”. That’s some pre-2025 buzzword lingo, clearly the model training corpus is out of date!
Currently I would not invest in any solution that enforces usage of a single LLM model provider, as the scene is really dynamic with the best player changing every month. Now Gemini 2.5 looks really great, but Anthropic, OpenAI, DeepSeek and others are probably working on the next, better model.
So, Aider + whatever seems to be working best, seems to be the best strategy that works for me. Aider has this nice feature that I can specify what files are added to the context, so I pay less and I do not confuse AI with stuff unrelated to a given task.
So, Aider + whatever seems to be working best, seems to be the best strategy that works for me. Aider has this nice feature that I can specify what files are added to the context, so I pay less and I do not confuse AI with stuff unrelated to a given task.
So much worse than Devin/Replit.
Also, the system prompt encourages it to use Genkit, so it tried to add AI to an email sending function...
Also, the system prompt encourages it to use Genkit, so it tried to add AI to an email sending function...
Have you seen gmail recently? They've plugged in ai all over it
It makes the most sense (to me) for app platforms like Firebase to invest deeply in the AI application development wave. Creating something from scratch with prompts is impressive but the gap between something that works locally to something that people can use is still large enough to require some dev & ops expertise.
v0.dev has bridged that gap pretty well. The deploy button usually just works and your app is then live on a vercel.app subdomain.
This looks very much like Replit. I like Replit, built a bunch of hobby projects mostly for myself, even though some of them are deployed and publicly accessible.
You will inevitably run into some doom loops, but we're still at the beginning. These things will probably be super powerful in 2 years time.
You will inevitably run into some doom loops, but we're still at the beginning. These things will probably be super powerful in 2 years time.
I’m a teacher, so different than most folks posting here. I have a little coding experience, but it’s definitely not my thing professionally. I just made several utility apps with this that I’ll be using in the classroom.
Asked FBS for a simple web form, it immediately went with NextJS, TailwindCSS and a whole bunch of build tools. Ask it again to change it to vanilla JavaScript, now it's stuck in an infinity loop... bad vibes : (
Can't we just have an AI that runs like a window manager? I.e. it has access to all apps on my desktop. And it can help me coding, but it can also help me write documents, draw stuff, edit photos and videos, etc.
That's the holy grail, my friend. Whoever succeeds will be the successor of 90s Microsoft (OS) and 00's Google (Internet/Search) in creating a unified platform which your grandmother associates as being "the computer". I fear that all the major players are trying exactly that (which is the reason why no one tries to be just a foundation model supplier). Some have the Hardware (Apple) but lack the software execution, others have the software (OpenAI) but lack the Hardware.
Well, you can start with a plastic cup, gain users and momentum, and work towards that holy grail. Makes sense that everybody wants that ultimate control over the user('s computer).
Apple seems to agree that that would be the ideal UX for AI, but for any computer, (1) it's going to require massive rework at the OS level, and (2) app developers have to be willing to integrate with the OS-level AI. I'm not sure that the latter is happening since profit incentives much better align with walled gardens.
Firebase for me is a synonym of running up large unexpected bill.
Is that possible?
Billings systems seems to be so opaque and complex.
and can it learn my private keys and then expose them to someone else?
Is that possible?
Billings systems seems to be so opaque and complex.
and can it learn my private keys and then expose them to someone else?
It's amazing.
I've wanted a small prompt-manager chrome extension for a while.
Was procrastinating.
Was able to build one for myself with Firebase studio in 30 mins.
Here's my PromptPal - built in just 30m (disable ad-blocker to avoid issues - there's some interference for some reason):
https://9000-idx-studio-1744253706406.cluster-fkltigo73ncaix...
No frustration whatsoever
Their prototyper is awesome
And code mode also great
I was able to push to github as well with no problems. And the tool generates nice commits for every single change one makes.
I've wanted a small prompt-manager chrome extension for a while.
Was procrastinating.
Was able to build one for myself with Firebase studio in 30 mins.
Here's my PromptPal - built in just 30m (disable ad-blocker to avoid issues - there's some interference for some reason):
https://9000-idx-studio-1744253706406.cluster-fkltigo73ncaix...
No frustration whatsoever
Their prototyper is awesome
And code mode also great
I was able to push to github as well with no problems. And the tool generates nice commits for every single change one makes.
I’m not sure I’ve ever felt my age more than reading “vibe coding” a half dozen times in this comment section.
I've been quite unimpressed with Google products for a while
> Out of capacity. Please try again later.
next level vendor lock in
Stopped reading after ai
- Create Angular app (success)
- Add SSR (fail)
- Give the correct command to add SSR (success)
- Add i18n (fail)
- Retries like 6 times, completely messes angular.json file up to no recovery, project doesn't even build
Outputs: It seems I am running circles, I cannot help you with that.
Wow, AI will really replace developers soon. /s
- Add SSR (fail)
- Give the correct command to add SSR (success)
- Add i18n (fail)
- Retries like 6 times, completely messes angular.json file up to no recovery, project doesn't even build
Outputs: It seems I am running circles, I cannot help you with that.
Wow, AI will really replace developers soon. /s
Similar experience, generated code confidently, manually had to copy paste it in places, only to find out not only it is bad but even the code was terribly structured / random gibberish sometimes.
Google just launched their vibe-coding product Firebase competes directly with Cursor, Lovable, Windsurf etc
I gave it the following initial prompt:
> An app where you input a question, then flip some coins to generate an I Ching prediction, and it generates a prediction / fortune for you. Then this combination of results can be fed to Gemini AI to produce a more detailed prediction text.
It generated something that looked fine. When I input a question and press the button nothing happened. After asking it to fix the problem multiple times and having it fail, I looked at the browser console to figure out the errors it was getting. Then I copied those errors and told it to fix them. After a few iterations, it solved every error and would generate a result. It completely forgot the part where you are supposed to flip coins before getting a hexagram to generate a fortune. After a bit of prompting, I was able to get it to display the hexagram and input question. However, sometimes it becomes confused about which hexagram was generated.
Overall, my impression is that these tools are still in the toy novelty stage rather than something you'd want to use for anything important.
Here is a screenshot of the app output for the question: Will Hacker News like my vibe coded oracle? [0] As you can see, it says that the generated hexagram is 24 or 41, but in the fortune text below it says 11.
[0] https://files.catbox.moe/i8t7rw.png