About BrainLab Games

The story

I'm Louis. I built the first version of this on a Saturday morning because my six-year-old daughter asked if she could make a video game and I tried the standard answers — ScratchJr, Roblox Studio, a tablet drawing app — and none of them worked for her. She's a creative kid. She's not a coder. None of those tools met her where she was.

I had the unfair advantage of being able to write code. So that afternoon I prototyped what eventually became BrainLab Games: kid types a sentence in plain English, AI builds a playable browser game, kid plays it, asks for changes, plays again. The first prompt my daughter typed was eight words long. The result landed in seventy seconds. She played it for an hour.

That afternoon was March 2026. I am writing this in May 2026 and she has made forty-three games. Her younger cousin, who is four, plays many of them. My friend's nine-year-old has built a parallax-scrolling platformer with a boss fight, unprompted by any adult. The thing the prototype was solving — the gap between a child's idea and the thing they wanted to play — turned out to be a real gap, and closing it turned out to matter.

What we believe

Articulation is itself a creative skill. The kids who learn to say what they want, edit until they get it, and iterate on the result are practising a higher-order creative act than the kids who learn to drag the right block onto the right shape. Both matter. We do the first.

Kids should see the result of their imagination, fast. The longer the loop from idea to playable, the more kids you lose along the way. Sixty seconds is the bar we set. We have not yet broken it.

Safe-by-architecture beats safe-by-policy. Most of the worst things that happen to kids on the internet happen on platforms that allow strangers to contact them. We don't. There is no chat. There are no public lobbies. Every share is a private link. We made this choice on day one and we won't change it.

Parents pay, kids play. Children are not the customer. Their parents are. That alignment means we never have to design for kid-retention at the expense of kid-wellbeing — no streaks, no daily-login rewards, no FOMO mechanics, no notifications. The kid can put it down and we don't care.

The model trains on outputs, not children. Children's prompts and their generated games are not used to train any model. We pay our AI provider for inference; that's the trade. The children produce no training signal and never will.

What we are not

We are not a coding curriculum. If you want your kid to learn to code, install ScratchJr (younger) or Scratch (older), or pay for a tutor. We are not a substitute for those.

We are not a babysitter. The kid still has to engage. If they don't enjoy the loop, no amount of clever prompting will make them.

We are not a Roblox alternative in the social sense. Kids using BrainLab Games cannot meet other kids through us. We think that's a feature.

The team

BrainLab Games is currently a small team. Most days that means me. We have part-time help with content and community. We do not have a sales team, a customer-success team, or a growth org. If you write to us, a person reads it and replies — usually me, within 24 hours.

If you want to reach me directly: [email protected]. I read everything.

Where we are

BrainLab Games is based in Singapore, but we serve families everywhere English is spoken — the US, the UK, Australia, Canada, New Zealand, Singapore, Malaysia. Our infrastructure runs on a Cloudflare Tunnel from a server I administer personally. Our email is self-hosted. We pay for one external service: our AI provider, who you can probably guess.

This is a deliberate choice. A small, comprehensible stack is easier to keep safe than a sprawling one. We can also keep prices low — $1.99 for a 10-game pack and $9.99/month for unlimited family use, with no ads ever, because we don't pay rent on a hundred SaaS vendors.

What's next

Things we know we'll build, roughly in this order: voice input (so pre-readers can really drive), better game sharing between siblings, a classroom tier, and longer-form games (a kid is currently capped at a few minutes of generation; we're going to push that).

Things we won't build: a chat surface, a public-creator economy, in-game purchases targeted at kids, NFTs, anything with the word metaverse.

If you're a parent reading this

Try a free game. It takes about sixty seconds and we don't ask for an email until you want to save what you made. If your kid loves it, we'd love to hear. If your kid bounces off it, we'd especially love to hear — that's where we learn.

— Louis