The two problems, separated
"I want to make and sell my own game" is actually two projects:
- Production — getting from idea to a thing people can play.
- Distribution — getting it in front of players, with a way to pay you.
The classic mistake is spending two years on #1 with no plan for #2, then discovering that launching into the void is the default outcome. In 2026 you can compress production to an afternoon (AI writes the code now — here's exactly how), which means you can afford to think about selling from day one.
Where you can sell a game, compared
| Channel | Cut they take | Upfront cost | Reality check |
|---|---|---|---|
| Steam | 30% | $100/game fee + you build it in an engine | Biggest audience, brutal discovery: ~19,000 games shipped there last year. You'll need marketing, a trailer, wishlists. |
| itch.io | You choose (10% default) | Free | Wonderful indie culture; pay-what-you-want norms mean most downloads are $0. |
| App stores | 15–30% | $25–$99/yr dev accounts + native builds | Millions of apps; paid downloads are nearly dead — you'd live off ads/IAP. |
| Your own site | ~3% (payments) | Hosting, storefront, taxes, everything | Total freedom, zero built-in traffic. |
| Arcade Sandbox | 25% — you keep 75% of every unlock | Building costs tokens (a 2D game typically ⚡350–650 ≈ $0.35–0.65); publishing is free | Browser arcade games only; first play is free, then players unlock your game for a flat ⚡500 ($0.50) — ⚡375 of it credited to you. |
None of these is "best" — they sell different things. Steam sells $15 premium experiences. Arcade Sandbox sells moments: quick, replayable browser games a stranger can be playing eight seconds after seeing your link. The economics match the scope: no fee to publish, first play free, a flat ⚡500 ($0.50) to unlock a game forever — and you keep 75% of every unlock (⚡375), credited to your token balance as build credit. It isn't cash out; it's fuel — player interest directly funds your next game. Building itself is priced transparently at what each job actually costs plus a margin, shown when the job finishes, so you always see where your money went.
The no-code path, start to first sale
- Build it (an afternoon). Type your idea into Arcade Sandbox, pick 2D (typically ⚡350–650 ≈ $0.35–0.65) or 3D (typically ⚡3,500–5,500 — frontier-model 3D genuinely costs more), and play the result two minutes later. Attach reference images if you have a look in mind.
- Tighten it (the important hour). Play it ten times. Fix what annoys you with plain-English revisions (typically ⚡150–250 each) — they stage as a private draft, so nothing goes live until you're proud of it. The bar: would a stranger play this twice?
- Publish (free, instant). One click puts it in the arcade with its own share page. First play is free for everyone — that's your demo. Unlocking unlimited plays costs a flat ⚡500 ($0.50), and ⚡375 of every unlock — 75% — is credited to your balance as build tokens.
- Share the link where your players are. A game link is uniquely shareable content: it's not a pitch, it's a toy. Discord servers, group chats, a subreddit where the theme fits, TikTok of the gameplay. Ten strangers playing beats a hundred impressions.
- Read the numbers, make game #2. Plays and unlocks tell you what landed. Your third game will be better than your first — the iteration loop is cheap enough to actually learn from.
▶ FROM IDEA TO ARCADE
Build this weekend's game now. Publish free. Keep 75% of every unlock.
Make your game →2D builds typically ⚡350–650 · 3D ⚡3,500–5,500 · revisions ⚡150–250 · $1 = ⚡1,000, tokens never expire · creators keep 75% of unlocks. · first token pack 25% off — $3.74
Honest expectations (read before quitting anything)
Will your first browser game pay rent? No. First games are how you find your voice; even the legendary solo devs shipped small and weird before they shipped hits. What a low-stakes arcade gives you is the full loop — make → ship → strangers play → tokens flow back — for pocket change, in a weekend. People who've run that loop ten times are the ones who eventually turn game-making into income. People still "learning the engine" in year two are not.
Three games that sell well in an arcade format
- The 30-second loop with a leaderboard. Score chasers get replayed; replays become unlocks. One strong verb, escalating speed, visible high score.
- The "one more run" roguelite-lite. Small random variation per run keeps a browser game alive for weeks.
- The gift game. Games made about something — a friend's birthday, an inside joke, a fandom — get shared into exactly the group that will unlock them. Cheap to make when a build typically costs ⚡350–650 (well under a dollar), weirdly effective.
Out of ideas? Pull the arm on CREATE1UP, our free idea slot machine, until something makes you grin — then build it before the grin fades.