Hitting a 5-figure MRR within a year by giving his AI coding tool away for
free
Less than a year ago, Eric Provencher built a free tool called Repo Prompt to scratch his own itch while building a game. Then, he monetized it, growing it to a 5-figure MRR. And yesterday, he left his full-time job to focus on building it out.
Here's Eric on how he did it. 👇
Taking the leap
I’ve been an XR developer for years, focusing on input and interaction research, specifically for hand tracking. I’ve got a patent in this space (granted last year!), and I published a cross-vendor OpenXR extension that moved the industry forward.
For the last year, I’ve been building Repo Prompt, largely through parental leave — my daughter is 9 months now! I just had my last day at work, though, so I’ll be spending a lot more time on it now, pushing ideas I’ve been thinking about for a while.
Repo Prompt is a tool designed for professionals tired of black-box prompt generation. It puts the emphasis on managing context for models, and is extremely efficient at applying edits directly in the built-in chat, or via a specialized XML format that models from any chat app can output. With that last mode, you can copy your prompt into ChatGPT, Claude, or AI Studio, and have the model output XML that can be pasted right back into the app to apply complex multi-file edits instantly.
It’s got a thriving community of power users and a 5-figure MRR!
Building Repo Prompt
I started Repo Prompt while working on a game hobby project for the Apple Vision Pro — it was a shared space game called Bomb Squad, that you could play ambiently while doing work or watching videos.
While it was a lot of fun to build, I was bottlenecked trying to use Claude Opus 3 on the Claude website, trying to make use of the 200k token context. Getting my files to the web UI and applying edits from it, was causing me immense grief and wasted time.
When Sonnet 3.5 came out, I realized how much faster I could build with it, and decided to prototype what was the first version of Repo Prompt, over a weekend using Electron!
All it did at first was apply filters, and let me select files from my repo, but it helped make me MUCH faster at prompting Sonnet, and I just dove down a deep rabbit hole trying to build out the next part of this problem — applying edits!
As I kept building, though, Electron was giving me a lot of grief, and I decided to restart and make this a native app, and that was a wonderful opportunity for me to learn Swift!
I make heavy use of both Appkit and Swiftui, though, more and more I have to switch important views over to Appkit for performance.
I don’t love that my tech stack limits me to only macOS users. I was surprised just how many folks on WIndows and Linux wanted to use this app, and I think the community of users could be far larger if I had support for them. But there really is a shortage of good native tooling for desktop.
And, as a solo dev, it would have been far more challenging to QA this app for so many different OS configurations. It just wouldn’t be as good for any platform as it is now for Mac users.
Testing and iterating
After about a month of work on the initial swift macOS native build, I had a pretty decent version working, and decided to try and find testers. I turned to Reddit! This was my first post about it! It had a google form sign up sheet, and at the time I manually added users to try it out.
Testflight was incredible early on. It took care of pushing updates to users, and let me focus on building, without even needing a website. There was a decent amount of interest, and a few started making their way into my discord.
Iterating for there was all about listening to user pain, and boy was there a lot of it!
Users were really into it, and they slowly kept finding the app from Reddit or word of mouth, so I kept building over nights and weekends - until my daughter was born in September.
While, at first, having a newborn takes all your time and energy, it also gives you a ton of time - as they nap A LOT! I spent many hours with a carrier as my baby napped, coding at my standing desk.
Communities and influencers
Early on, Reddit was the biggest source of new users, but I had to post so much that I was afraid I was becoming spammy, even though the app was free at the time. I was confident in my product, though, so I kept evangelizing and talking to anyone who’d listen.
From there, I started cold DMing folks on Twitter who were making AI coding content. I was hoping one of them would like the app enough to make a tutorial about it. The only one to really get back to me was Ray Fernando and he invited me to join his livestream, which ended up being one of his biggest episodes to date!
Still, though, after a couple months, growth wasn’t that significant, and I nearly gave up, not knowing where to go next with it, or if any more work would even be worth doing.
Around December, though, o1 Pro came out and, while it was absolutely the most powerful coding model that existed at the time, the tools to use it were just awful — unless you used Repo Prompt to build your prompts for it! Which is what convinced McKay Wrigley to give it a try.
He made a video about his workflow with it that put the app on the map and truly transformed my prospects working on this thing.
Free vs paid
As I kept building, the app kept growing. I had thousands of passionate, vocal users who were very engaged and truly helped shape the product. I had a donations page, and had made about $5k from that, which was decent signal, but it wasn't enough money to commit so much of my time.
Around that time, I had the first potential investor come around, and I needed to make a choice on a path — was I going to pursue VC for this, or just stay bootstrapped. I needed to know if users would pay for this thing, in order to make it sustainable. I was terrified of finding out. Around late February, I started grinding on my landing page and mulled over pricing.
Was I going to do monthly or lifetime? A lot of folks advised me to set up a monthly sub, but some vocal users pleaded with me to do lifetime.
On March 9th, 2024, I went live with the lifetime offer and gave significant discounts to some folks who were passionate early testers and to those who donated, which made all the difference.
To say that some testers were upset is an understatement. How dare I charge for this tool they’d been using for free! I had a free tier still, but the limits were much too low, they said.
After talking to users, and seeking the advice of friends, I doubled the free tier limits, and added a monthly billing option. It wasn’t perfect, but the most vocal criticisms died down and have not re-emerged since!
After that, the community actually blossomed, and this app became a sustainable project. I really love all the folks in the Repo Prompt discord. They’re some of the most brilliant frontier AI coders around, and I learn from them every day!
My advice to anyone going through a similar transition, is to talk to your users, listen to them, but also trust yourself and do what is best for you and your business. No one wants to pay money for things they like to use, but at the same time, people don’t always value things that are free!
Build community and listen
Here's my advice. Focus on building community around your product, first and foremost. Word of mouth will be the most powerful sales tool for any product you make.
Follow your curiosity. For me, building this app was mostly a matter of following my curiosity, and I think that was super essential for me getting to where I am today.
And make sure that you’re engaging with users and listening to them!
What's next?
Over the coming summer I want to revitalize Repo Prompt and focus it as a way to integrate with all coding agents via MCP. I want to turn into a powerful way to steer them, and enhance them with the insights of external powerful models.
The AI coding space is rapidly evolving, overall, and being able to stay relevant with a feature set remains an ongoing challenge, as competitors are extremely well funded, always pushing the envelope.
You can follow along on X. And check out Repo Prompt!
Leave a Comment
Giving it for free? Tell me how do I can market like I did organic marketing on Tiktok and no one is seeing or interacting with my videos. How can I get users and have conversion rates?
'Tis the challenge of marketing. Developing something both eye-catching and effective - and reaching the correct audience. It takes time.
Thanks!
This is such an inspiring journey — the blend of raw technical drive, community listening, and sheer persistence really comes through. It’s rare to see someone not just identify a personal workflow pain point, but turn it into a tool that resonates so deeply with others. Also love how you embraced constraints (like macOS-only) while still staying focused on quality. The part about coding with your newborn in a carrier is especially touching — reminds us that great software often has very human roots. Thanks for sharing all this so candidly.
Congrats on your success James, and particularly the new addition to your family!
I’ve had a similar experience with moving from free to paid. An initial minor backlash via email (particularly via those that “donated”) but it quickly quieted down. Haven’t gotten a grumpy email in months now that people are used to it being a paid service.
To mollify the donators I proactively offered a credit of double their donation for anyone that contributed before I put up the pay wall, provided they subscribed. This helped get early signups, and makes it easier to apply account credit on the Stripe customer vs managing it in my own system.
I also offered to refund their donation if they wanted - exactly zero people asked for a refund.
My advice to others is to monetize as soon as you start feeling product-market-fit begin to pull. You’ll know it when you see it. The longer you wait the more people get used to “free”. Charge on the higher end of what you’re comfortable with and that is still probably too low 😜 We tend to be our own biggest obstacles when it comes to monetización and pricing. Just go for it.
This was really inspiring to read! I'm planning on starting something similar to this. I just created my v1 product, and am wondering how you handled any user feedback that felt discouraging or unclear?
it's amazing
Great story!
Would love to hear what works for you from marketing perspective, and how specifically you started to market it. Thanks.
its my dream too, building my saas and quit my full-time job
Great content
Coollll
This is awesome — what tech stack did you use?
I think it's a great idea to give away some tools for free.
This is basically gathering trust, and ensuring your audience will come back again for other tools!
It's also helping your audience by providing valuable tools for them to grow.
How do you balance free features vs. premium to avoid cannibalising revenue?
I think pricing is a very important factor.. you don't want wo loose your community right away, but also have to be sure to get enough money to get it going. Also a point to consider is word to mouth, which is very important for this kind of "app".. but i think with the double amount for free tier and the monthly billing its very fair for anyone.
Absolutely loved reading this. The balance of parenting, coding with a baby in a carrier, shipping, and building a real community is seriously inspiring. The product evolution, from solving your own workflow pain to hitting meaningful revenue, is such a strong example of following user need over hype.
I’m building something design-focused right now, and your shift from technical curiosity to sustainable product really resonated. Bookmarking this for those inevitable “should I keep going?” days. Massive respect 👏
marketing is one of the most diverse sectors - you can even give away your product at a loss and find a way to profit largely off of it like Eric.
Great story! That's awesome.
Selling saas is the way
It’s very interesting, but there are many aspects that will require much deeper and longer analysis.
In less than a year, he achieved 5-figure MRR by offering his AI coding tool for free, attracting users quickly, and then monetizing premium features and services as demand and trust grew.
hello people
Love this journey! The "scratch your own itch while on parental leave" origin story is so relatable. Building while baby naps = ultimate productivity hack 😅
Key insights:
Free → community → paid worked perfectly here. The fear of monetizing is real, but users who truly value your product will pay
McKay Wrigley's video was the inflection point - shows how one good influencer mention can transform everything
Native Mac focus was smart - better to nail one platform than spread thin across all
The transition from donations ($5k) to 5-figure MRR is impressive. Lifetime vs monthly pricing is always a tough call, but sounds like giving users both options was the right move.
Question: How are you planning to differentiate as the AI coding space gets more crowded? The MCP integration angle sounds promising for staying relevant.
Congrats on taking the leap - excited to see where Repo Prompt goes next!
"No one wants to pay money for things they like to use, but at the same time, people don’t always value things that are free!"
Simple but true!
I'm new to this world, joined IndieHackers in attempt to see if anyone had a co-passion in developing an app with me that i think could have a large impact on humanity. I've stumbled upon this thread, and while i'm not familiar with many of the processes and lingo used, it's inspiring to see the perseverance and passion that resonates through your profile.
By offering his AI coding tool for free, Fleischmann attracted users quickly before monetizing through premium features and developer-focused upgrades.
Thanks for sharing your journey, Eric! It’s inspiring to see how dedication and persistence can turn an idea into a successful product , especially while balancing family life.
I really like your emphasis on building a community first; it’s something I’ve seen make a huge difference in SaaS growth. Listening to users while staying true to your vision is definitely a challenging balance.
If you don’t mind me asking, how do you decide which user feedback to prioritize when planning new features? Looking forward to following your progress and learning more!
By offering his AI coding tool for free, James Fleischmann achieved five-figure MRR in just one year, demonstrating the power of value-driven growth and smart monetization.
Love the 'give it away free' strategy! This resonates deeply with our approach at AlbumForge.
We do 1=1 giving (every license sold = 1 donated to kids in need) and built our AI co-creator Claude with ethical principles.
Your story proves that generosity can be profitable. We're seeing similar results - when you lead with value and purpose, the revenue follows naturally.
How did you balance free vs paid features? And did the giving strategy help with user acquisition beyond just retention?
Built trust by offering real value upfront a strategy we also believe in at AIChief, where we feature powerful AI tools that start free but scale fast.
I can't believe you developed while on paternity leave. I'm also on parental leave and I admire you.
This was really inspiring to read! I'm planning on starting something similar to this. I just created my v1 product, and am wondering how you handled any user feedback that felt discouraging or unclear?
You just forgot to give tons of credits to the mom that took care of your baby while you were working on your project :)
crazy
This is a great story about the power of scratching your own itch. It's interesting how you frame the native macOS choice as a limitation. An alternative view is that it's your biggest strategic moat. While web-based competitors compete horizontally on features, you're competing vertically on performance and deep OS integration. That focus builds a stickier product and a more loyal user base that's harder for cross-platform tools to replicate.
Congrats on the impressive growth! What's your strategy for converting free users to paid subscribers? Is there a specific feature or usage limit that triggers the upgrade?
The main strategy for converting users revolves around usage limits. Initially, the app was completely free, which was great for building the community but wasn't sustainable long-term. When I first introduced the paid version, the trigger for the upgrade was hitting those limits. To be honest, I set them too low at first and got some very direct feedback from early users. After listening to them, I doubled the free tier limits. So the core strategy is to find a balance where the free tier is genuinely useful for casual use, but power users who rely on the tool for their daily workflow see clear value in the paid plan. It's less about a single "magic" feature and more about letting dedicated users hit a fair usage ceiling that encourages them to support the project's continued development.
Hope this help.
Glad you're not working for a company/government that would punish you for spending your paternity leave on production! I can say for those in Canada, we are sincerely jealous of your opportunity - congratulations!
damn that's smart
Congratulations to you for having succeeded, I've been here in the fight for a while now and I can't. Could you give me some basic advice so I don't give up? Because that's all that comes to my mind.
This is super timely. I’m also building something in the compliance/legal tooling space — and it’s wild how confusing GDPR/CCPA requirements still are for small teams. Curious: how are you keeping up with changing regulations across regions? Manual research or legal API sources?
Круто !!!
If you're seeking a trustworthy and effective credit repair service, I wholeheartedly recommend Knight Lightning Credit Repair. Their team is dedicated to helping clients improve their credit scores and achieve financial stability. What sets Knight Lightning Credit Repair apart is their personalized approach. They take the time to understand each client's unique financial situation and tailor their strategies accordingly. Their commitment to transparency and ethical practices ensures that clients are well-informed and confident throughout the credit repair process. Clients have reported significant improvements in their credit scores, often within a few months of working with Knight Lightning Credit Repair. Their expertise in navigating credit bureaus and disputing inaccuracies has proven invaluable for many. For anyone looking to rebuild their credit and take control of their financial future, Knight Lightning Credit Repair is an excellent choice. You can reach them at via 609 616-2261 to learn more about their services and how they can assist you.
Wow, it’s like watching a phoenix rise from the ashes! Gave away a tool for free and now it’s cashing in five figures—talk about a win!
Huge congrats on the leap — Repo Prompt sounds like a powerful and much-needed tool. Curious if you’re planning support for more model-specific features down the line? Excited to see where you take it next!
how did you get users, people now-a-day even did not come for free products
Really appreciate your experience and effort!
It's definite a tough path, but you've created you own story.
This is such an inspiring journey! Starting with a free tool, listening deeply to user pain, and turning it into a sustainable, paid product shows the power of community-first growth. Your story really resonates at Bestdesign2hub we also believe in building trust before monetizing. Thanks for sharing the behind-the-scenes with such honesty
this is very helpful and amazing journey
This was very helpful, especially where you shared your experience discussing Repo Prompt in relevant communities as well as the influencers that brought you early success. These anecdotes are invaluable to new founders to learn how to bring attention to their products and services. I hope you have continued success with your venture!
This was a great read. I really respect how you followed your own pain points and curiosity to build something useful and stuck with it through all the ups and downs. The honesty around pricing and platform tradeoffs hit home. Excited to see where Repo Prompt goes next, especially with agents in the mix. Cheering you on!
Incredible journey—loved how organically Repo Prompt emerged from solving your own frustrations. This is a powerful reminder: the best products often start by scratching your own itch. Big congrats on taking the leap and building a vibrant community around it! Excited to see where you take it next 🚀
Really liked your take—especially the part about how free things aren’t always valued. That mindset shift is key. I’m building something too, and talking to users early has made a huge difference. Excited to see where Repo Prompt goes!
Really appreciate this perspective — especially the part about not always valuing what's free. That mindset shift is so important when building something sustainable.
Also completely agree on the power of building community early. I’ve been working on CompliAssistant, an AI-powered HIPAA compliance tool for small businesses, and user conversations have shaped nearly every decision so far.
Following your curiosity while still engaging deeply with users feels like the sweet spot. Looking forward to seeing how Repo Prompt evolves — sounds like you’re tackling a really exciting corner of the AI space!
This comment was deleted 4 months ago