The best consumer founders know the game. 35K+ of them read Consumer Startups every week.
Stay ahead. Get the playbook behind today's breakout startups.
Read time: 3 mins 22 seconds
Four years ago, two brothers made a bet that sounded a little crazy. That AI would one day let anyone build a real video game just by describing it.
Amir and Ali Sadeghian started Astrocade in 2022, before ChatGPT existed, and spent years building toward that bet with the godmother of modern AI, Fei-Fei Li.
Then it clicked. 8 months after launch, Astrocade has 5M monthly users, 140M game plays, and a fresh $56M from Sequoia last month.
Excited to share the playbook behind one of the biggest consumer AI breakouts of the year 👇.
The bet before ChatGPT
Two Iranian Americans, Amir and his brother Ali, co-founded the company in late 2022.
Stable Diffusion had just come out, and ChatGPT didn’t exist yet.
The brothers moved around constantly as kids, country to country. Each new place came with new adventures and new people, and that feeling of discovering a new world became the founding inspiration: let anyone build and share their own interactive world.
The founding team was stacked. Amir came from Stanford's AI Lab. Ali was ex-Google Research. They were joined by Fei-Fei Li, the godmother of modern AI and co-director of Stanford's Human-Centered AI Institute.
Ali likes to explain the market with a classroom question. Ask a classroom who plays games, and everyone raises their hand. Ask who wants to create a game, and 80% say yes. Ask who has actually made one, and you get one or two hands.
Everyone wants to make games. Almost nobody can.
The brothers bet that AI would eventually close that gap, and they just had to survive until the models caught up.
Astrocade’s MVP
Their MVP, shipped in early 2023, was template-based game creation for platformer and side-view games. AI back then simply couldn't generate full games, so the team designed around the limitation instead of fighting it.
This constraint forced them to be creative, and it delivered their first proof that the idea had legs.
They had two key hypotheses to validate:
Would people actually make games with AI?
Would those games be good enough that players genuinely enjoy them?
Both came back as yes. The creator community grew, and they also started seeing player growth along with higher-quality engagement.
Building TikTok for games
1/ The new Astrocade
In July 2025, Astrocade shipped a near-complete product overhaul.
Three things changed:
AI brainstorming flow. Instead of dropping users in front of a blank text box, the app starts with conversational ideation that turns a half-formed idea into a full game blueprint. This solved the biggest barrier for new creators: the blank page problem.
Game-specific editors. Instead of one generic editor, Astrocade's AI assembles a custom toolkit for each game. A racing game gets different editing tools than a platformer, which makes iteration much easier for non-technical creators.
One-tap remixing. Every public game can be cloned, tweaked, and re-shared. Users don't have to invent from scratch. They can riff on games that are already working.
Under the hood, a team of specialized AI agents collaborates in real time on art, UI, narrative, audio, and mechanics.
The remix feature turned out to be especially important. Every game on the platform became a starting point for the next creator, which meant the content library started compounding on itself.
Astrocade publicly opened to the world in August 2025 with the new product. The team was 13 people, and they had been building for almost 3 years.
Monthly game plays went from 0 at launch to 6M by late fall, then to 140M by spring.
Most of their best users are women between 20 and 40. Different from those hard-core gamers on Steam, Astrocade’s users are spending leisure time on Astrocade the way they'd otherwise spend it on Instagram.
David Cahn at Sequoia had known the brothers since 2022 and told them he'd invest as soon as they hit PMF. At a dinner in November, Amir told him the numbers were inflecting. At lunch shortly after, Amir turned his laptop around and showed him the Amplitude dashboard with strong engagement data across the board, paired with rapid user growth.
Cahn says they didn't eat a bite. Amir presented to the Sequoia partnership that same afternoon, and the term sheet was issued the next day.
2/ Astrocade’s creator-led growth flywheel
The biggest growth unlock comes from game creators.
Creators make games and share them. Players play those games and share them. Some of those players become creators themselves, and the cycle repeats.
A few things make this flywheel spin faster on Astrocade than on traditional UGC platforms:
AI removes the production cost, so creators can ship new games constantly
Remixing lowers the creation bar even further, since every game is a starting point for the next one
Creator monetization closes the loop. Top creators are already making thousands of dollars a month
They recently started a $10M creator fund to incentivize creators to create more quality games on the platform.
3/ The non-obvious hard parts
There are several non-obvious challenges in this space according to the team.
The AI creation tool itself. The quality bar is high and constantly rising. What was good enough last quarter is not good enough now.
The creator community. The team is emphatic that creators are what make the platform special and the games good.
The recommendation system. With 75K+ games on the platform, surfacing the right game to the right player is both crucial and hard.
UX for a brand new category. There's no design pattern library for an AI game creation social platform. Every interface decision is a first.
What stood out to me is that none of these are about model access, compute, or capital. They’re all classic consumer product problems.
🌶️ My take on Astrocade
The AI vibe app building market is heating up, with horizontal players like Lovable and Replit and vertical players like Astrocade jumping in to own this new creation paradigm.
One of Astrocade’s biggest direct competitors is Sekai, which I covered a few months back. It pivoted directly into vibe game creation and has seen over 15M games created in just a few months.
I wonder whether this market consolidates the way attention platforms tend to. Creators chase the biggest audience, audiences chase the best content, and the recommendation engine compounds with every play. Those loops have historically crowned one winner (e.g. TikTok, YouTube, etc). However, games are more varied and more remixable than video, so there may be room for several platforms to each own a corner.
Another thought is the impact on the gaming industry overall. Today this is casual play, closer to scrolling Instagram than booting up Steam. As the models improve, the obvious next step is deeper, more ambitious games. If that happens, traditional studios have a real problem at the low and middle end. I do not think craft disappears, but a lot of the market that studios consider safe could get generated on demand for free.
Zoom out and Astrocade is one instance of a pattern I am seeing everywhere. Ali's line that everyone wants to create something and almost nobody can is not really about games. It is about software. Everyone wants a custom version, and AI is finally cheap enough to give it to them. I recently saw something similar in travel booking, where a tool like Odessia behaves like a personal travel agent instead of a booking site, planning and booking a trip shaped entirely around how you travel.
We are shifting from browsing what already exists to generating exactly what we want.
BUT this raises an uncomfortable question. If the cost of creating anything drops to zero, where does the value go? Probably not to the creation itself. It likely flows to taste, to community, and to whoever owns the moment of discovery. Astrocade figured that out early. Most companies still think they're selling the thing they make.
What do you all think? Reply to this email. Would love to hear from you :)
See you next Tuesday,
Leo
Thoughts on today's email?
📥️ Want to advertise in Consumer Startups? Learn more.