I built an AI IDE alone in 6 months — it scores 26/26 vs Cursor's 10/26
Hey everyone
👋 I'm Eric, solo developer based in Paris. For the last 6 months, I've been building Dumble — an AI-powered IDE that lets anyone create websites, apps, and documents without writing a single line of code. WHY I BUILT THIS I got frustrated with Cursor, Bolt, and Lovable. They're impressive but expensive, limited, and they forget everything between sessions. Every time you start a conversation, you're a stranger again. So I decided to build my own. WHAT MAKES DUMBLE DIFFERENT
🧠 Permanent memory — Dumble remembers your projects, preferences, and context forever
🤖 Autopilot mode — describe your project, Dumble builds everything automatically 🎙️ Voice mode — talk to your IDE like a colleague
📸 Real-time screenshots — it sees what it builds and fixes issues instantly
💰 Stripe integration — e-commerce ready out of the box
🚀 One-click deployment — your site goes live in seconds THE NUMBERS In an independent benchmark across 26 criteria, Dumble scored 26/26. Cursor scored 10/26. Bolt scored 8/26.
I'm a solo founder with zero funding going against companies worth $60 billion combined. No team, no investors, no marketing budget. Just me and my code.
REVENUE MODEL Freemium — free tier + paid subscriptions starting at €29/month.
WHAT'S NEXT I'm looking for early adopters and honest feedback. If you've ever been frustrated with AI coding tools, give Dumble a try and tell me what you think.
🔗 https://dumble.fr
AMA — happy to answer anything about the tech, the journey, or the business model.
The interesting part here isn’t just the IDE itself - it’s the positioning against Cursor/Bolt/Lovable.
More and more users seem to decide at the comparison stage now (“X vs Y”, “best alternative to…”).
That’s actually why I started building AIToolsRecap around reviews + comparison discovery instead of just tool listings.
Curious to see how Dumble evolves in this category 👍
It is an ambitious move to challenge established giants like Cursor and Bolt especially when the primary pain point for most developers is the "context window reset" that forces them to explain their project architecture every single morning.
The real differentiator here is the "permanent memory" because by maintaining long-term project context the AI can suggest modifications that align with your existing design patterns rather than introducing conflicting code that requires manual cleanup later.
Since you are running this as a solo founder with zero funding how are you handling the infrastructure costs of real-time screenshot analysis and long-term memory storage while still keeping your pricing competitive with VC-backed competitors?
Impressive solo build — permanent memory across sessions is a real differentiator, that's one of the biggest frustrations with Cursor.
One question,the 26/26 benchmark — who ran it and what were the criteria?
Hey Eric 👋 really impressive build — especially as a solo founder going up against big players, that’s not easy
The permanent memory and autopilot mode stand out a lot — that’s a strong differentiation from tools like Cursor
Since you’re currently looking for early adopters, this is the perfect stage to bring in targeted users quickly and get real feedback
We help early-stage products get their first traction through creator-led promotion across platforms like YouTube, TikTok, and Instagram
If you’re open, I can share a few ideas on how you can start getting consistent early users 🚀
Thanks for the kind words! Right now I'm focused on organic growth and building the product. I'll keep your offer in mind for later when I'm ready to scale marketing. 🙏Hey totally understand 👌 organic growth is actually the best foundation
If you ever want, I can just share a few non-sales growth ideas that worked for similar early-stage tools — no pitch, just insights 🚀
Either way, wish you the best with the build 👍
merci
You're very welcome! If you ever need any additional ideas or support, feel free to reach out. I’m here to help whenever you're ready!