Inside PeachJam

Why I’m sunsetting PeachJam

PeachJam didn’t work as a business. I’m closing the builder, keeping this page, and moving on to something more ambitious.

I built PeachJam around a simple bet: people wanted something like Lovable, but cheaper. There was real engineering behind it. You could describe an app, iterate in chat, preview it, and publish it. We built templates, backends, and a lot of the machinery around that experience.

But making the product more complete didn’t make the bet more convincing. Here is where I think I went wrong.

“Cheaper” wasn’t enough

My comparison was PeachJam at $5 against Lovable at $25. I thought that gap would give people a good reason to try us and switch.

What I underestimated was the comparison people were already making. If a tool saves engineering time, $25 can already feel extremely cheap. Saving another $20 is much less compelling than saving hours or days of work. Trying an unfamiliar product also costs attention, trust, and time.

I had a lower price. I didn’t have a sufficiently different reason to choose PeachJam. If price is your main advantage, an incumbent lowering its price can erase it. Lovable’s costs might have made that difficult, but hoping a competitor cannot afford to undercut you is a weak foundation.

I stopped believing in the destination

As I used Codex and Claude Code more, I became less convinced that a dedicated “build websites through chat” product was the right place to spend my time. In my own work, the general coding agents felt more capable, more generous, and useful across a much wider range of tasks.

The clearest signal was personal: I started using Codex, and I couldn’t get myself to use PeachJam enough. The product I was building wasn’t the product I wanted to reach for.

My bet is that the gap between specialist app builders and general coding agents will keep shrinking. I don’t believe Lovable’s market, in its current form, will last for very long. I could imagine Lovable being acquired by an AI lab or a larger company; Google would be my guess. That is speculation, not inside information.

I could be wrong about that future. But I no longer believed that chasing Lovable was ambitious enough to justify the work ahead. It is hard to build a good company around a destination you have stopped wanting to reach.

I chased feature parity before testing demand

I should have shipped a small MVP and started marketing immediately. Instead, I tried to get close to Lovable’s feature set first.

The scope kept expanding. So did the technical problems. Marketing started late, after a lot of engineering effort had already gone into a product whose central differentiation was still price.

Some of that work made PeachJam better. It did not answer the question I needed to answer first: would people care enough about this particular product to use it instead of what they already had?

The lesson I’m taking forward is to put the core promise in front of people much earlier. More features cannot substitute for a reason to switch.

What happens to PeachJam

The builder is being retired. I’m preserving the project data and keeping a small public archive of the landing page. I’m also working to keep the projects I personally shared online available where their dependencies make that practical.

Those projects deserve to stand on their own. Keeping them accessible should not require keeping an entire app-building business running behind them.

What I’m building next

The next project is Singularity, previously called Circle. The ambition is a full SaaS company builder: you set the vision and the budget, and AI agents handle the continuous work of improving the business.

That means engineering and fixing the product, marketing and posting, and using revenue and conversion results to decide what to try next. A working loop from an idea to a change, to a measured result, to the next decision.

My bet is that the capabilities needed for this will become practical over the next two or three months. That is a thesis I need to test, not a promise that the technology or the business will work on schedule. I want to use that window to market, implement, and ship alongside the capability.

This time, I want the marketing and the product to develop together. I want to find out whether people want the outcome before I disappear into building the entire machine.


Thank you to everyone who tried PeachJam, shared something they made, or told me what wasn’t working. I’m proud of parts of what we built. I’m also ready to be honest about why it wasn’t enough.

Tornike

Follow what I’m building next