I’m Adi, co-founder of MXFLO. We are building an iOS/Android app that lets anyone create, play, and monetize games entirely on their phone. We are building Replit meets Roblox for games, but mobile-first.
Users use our simple, chat-based editor to vibe code games in minutes, with no PC or coding experience needed. Games are instantly playable and published to a "TikTok-like" discovery feed where they can go viral.
This started with a personal frustration. I love playing simple games like Wordle before sleep, but I hated paying $6 a month for the NYT Games app. It felt crazy that an app hosting simple word games makes over $100M a year, while basic LLMs can now generate infinite variations of these games for free.
At the same time, you have Roblox making $4B a year. They proved people want to create, but they only care about the 1% of users who have a PC and can learn complex desktop tools. They treat the other billions of mobile users just as passive consumers.
We think that is a massive missed opportunity. We want to do for gaming what YouTube/TikTok did for video. We want to unlock that 99% of mobile-only users and turn them from players into creators.
What we are building:
1. A "vibe coding" interface. You describe your game in chat, and our backend (using Claude Sonnet 4.5 and Stable Diffusion XL) generates the initial game loop.
2. Instant Play. The game runs immediately in a secure WebView on your device. No compiling, no waiting.
3. On-Device AI (The hard part). Right now we do inference in the cloud. Our long-term technical moat is moving 2D asset generation directly onto the phone to reduce latency and server costs. My co-founder Tejas has been working on edge AI optimization since 2017 and it is a massive technical challenge we are excited to solve.
We are currently in closed beta and applying to YC. I put up this landing page to see if others agree that this is a problem worth solving.
I would love to hear from this community:
- Do you think a "mobile-only" workflow can ever be powerful enough for real games?
- What is the biggest technical bottleneck you see for on-device game generation?
adithiya_shiva•2mo ago
I’m Adi, co-founder of MXFLO. We are building an iOS/Android app that lets anyone create, play, and monetize games entirely on their phone. We are building Replit meets Roblox for games, but mobile-first.
Users use our simple, chat-based editor to vibe code games in minutes, with no PC or coding experience needed. Games are instantly playable and published to a "TikTok-like" discovery feed where they can go viral.
This started with a personal frustration. I love playing simple games like Wordle before sleep, but I hated paying $6 a month for the NYT Games app. It felt crazy that an app hosting simple word games makes over $100M a year, while basic LLMs can now generate infinite variations of these games for free.
At the same time, you have Roblox making $4B a year. They proved people want to create, but they only care about the 1% of users who have a PC and can learn complex desktop tools. They treat the other billions of mobile users just as passive consumers.
We think that is a massive missed opportunity. We want to do for gaming what YouTube/TikTok did for video. We want to unlock that 99% of mobile-only users and turn them from players into creators.
What we are building:
1. A "vibe coding" interface. You describe your game in chat, and our backend (using Claude Sonnet 4.5 and Stable Diffusion XL) generates the initial game loop.
2. Instant Play. The game runs immediately in a secure WebView on your device. No compiling, no waiting.
3. On-Device AI (The hard part). Right now we do inference in the cloud. Our long-term technical moat is moving 2D asset generation directly onto the phone to reduce latency and server costs. My co-founder Tejas has been working on edge AI optimization since 2017 and it is a massive technical challenge we are excited to solve.
We are currently in closed beta and applying to YC. I put up this landing page to see if others agree that this is a problem worth solving.
I would love to hear from this community:
- Do you think a "mobile-only" workflow can ever be powerful enough for real games?
- What is the biggest technical bottleneck you see for on-device game generation?
Thanks!
Shiva_68•2mo ago