This is how we're surviving against funded competitors for our vibe coding platform.
Me and my co-founder are building vibe coding platform for mobile apps. Where you can build mobile apps with supabase as backend along with AI capabilities or just UI depending on your requirement.
It all started with Paul Graham's advice.
We launched FlexApp publicly on March 15th after our first demo went viral with 200 users on waitlist.
We were on the right path: a niche tool to only build mobile apps.
Soon hitting 50 new signups daily.
All these signups came with massive LLM costs. Starting with just ~$4K, we burned through 30% of our budget in the first week with only $260 MRR.
Our well-funded competitors offered 5 free code gen requests daily (30/month), while we provided just 1 request/user lifetime.
Talked to users, they wanted to explore platform for free but even with our minimal free tier, costs were eating up our savings. Traffic kept growing from our use case videos, engagement on Reddit/X and blogs. We approached VCs but got no replies.
We pivoted to a paid-only model after burning 60% of our funds. Still getting ~50 daily signups from our @daily.dev blog, but users expected the same free plans as competitors. Applied for credits but never got it approved. We decided to approach VCs again—this time for solutions, not money.
We found @paulg's email on the website and decided to write an email asking for advice. And he replied with the below, here's the link.
"Start with a small, specialized subset of the market whose inhabitants will pay you. Do consulting if necessary, using your own app."
We approached indies like @nico_jeannen , Marc Louvion, @dannypostmaa and SMBs.
We decided to offer services where we would build mobile apps for them using our own platform in return for shoutouts/feedbacks.
Then for SMBs, we wrote personalised emails. Every emails included process, benefits and timeline.
Key point : It has to be personalised, and short.
Luckily, within 2 weeks we got 2 customers(1 from DM and 1 from email) paying in total $3.8k.
We will build it with our own platform and meanwhile it will help us to know our product from the user's point of view.
We will use this opportunity to make our product even better for all.
We will soon launch FlexApp on product hunt as well.
Any suggestions or advices are welcome. :)
Replies
This is one of the most honest founder posts I’ve read in a while.
Burning 60% of runway because of LLM costs is the part most AI builders don’t talk about publicly. Distribution is easy when demos go viral. Unit economics is where reality hits.
The Paul Graham advice you quoted is timeless for a reason:
What you’re doing now is actually very strategic:
• You validate willingness to pay
• You reduce LLM burn (paid projects justify cost)
• You dogfood your own platform
• You collect high-quality feedback instead of random free-tier noise
That’s not a pivot. That’s tightening focus.
From our experience at Mobiwolf building mobile products (and now growing our blockchain & AI-driven projects), service + product hybrid models are underrated. Especially in early stages. Services fund product development. Product improves service margins. It becomes a flywheel if executed carefully.
A few thoughts:
1️⃣ Be extremely opinionated about who FlexApp is for.
“Vibe coding for mobile apps” is cool — but who pays fastest?
Indie hackers? Agencies? Non-technical founders? SMBs?
2️⃣ Consider usage-based pricing instead of hard paywalls.
Maybe limited AI credits but full UI builder access?
Lower friction to try, but predictable cost ceiling for you.
3️⃣ Turn your service clients into public case studies.
“Built in 7 days with FlexApp”
That narrative can be stronger than feature lists.
4️⃣ Track LLM cost per generated screen / feature.
If you can show: “Average app built for $X infra cost” — that’s investor gold.
Also — launching on Product Hunt after you have paying users and real case studies will hit differently than launching with just a viral demo.
Big respect for not giving up after the burn. That phase filters out most founders.
Curious — are your two paying clients building internal tools, consumer apps, or revenue-generating apps? That insight might shape your niche positioning massively.