Gem Match Template: A Match-3 Game You Can Reshape

Gem Match Template: A Match-3 Game You Can Reshape

Gem Match is one of the five game templates in Zugo: a match-3 puzzle on canvas with real cascades, a rising combo multiplier, a move counter and a score goal that raises itself when you reach it. The clone costs no credits, so the first playable version is free, and everything after that is chat edits.

What follows is the mechanic as it is actually coded, the controls, which changes take one sentence and which take three, and the honest edge of what a browser game template can be.

What kind of game is Gem Match?

It is the classic version of the genre. You swap two adjacent gems to line up three or more of a kind, matched gems vanish, the ones above fall into the gap, and if the fall creates a new line the cascade continues with a growing multiplier.

The board is eight by eight and there are six gem types. Both numbers matter more than they look. Fewer types makes the game trivial, more turns the board into noise where no move is readable at a glance, and the board size sets how many legal moves exist at any moment.

Play is limited by moves rather than a timer, which makes it a calm game: you can think as long as you like. A run starts with twenty five moves and a score goal, and reaching the goal refills moves and raises the target, so difficulty climbs by demand rather than by clock.

The screen shows four things: score, best, moves remaining and the combo multiplier. There is a reshuffle for dead boards, a hint that appears after a few idle seconds, and a title screen that plays the game to itself when nobody is touching it. That attract mode is worth keeping through any redesign, because a still picture explains match-3 far worse than the game explaining itself in the background.

How do you play it, and on what?

The controls are doubled deliberately so the same file works with a mouse, a finger and a keyboard.

The swipe is the one to test properly. Match-3 is played on phones far more than on laptops, and on a narrow screen the cells get small, so what feels precise with a mouse can miss with a thumb. Publish and open it on an actual phone before you decide the game is finished.

The reshuffle deserves the same attention. It is the escape from a dead board, and if you change the board rules with edits, check that it still works. A puzzle with no way out of a stuck position gets closed and never reopened.

Nothing explains the rules in words, and that is correct. Match-3 teaches itself with the first swap, and an instruction screen only delays the moment the player opened the game for.

What can chat edits actually change?

Games are edited with the same plain sentences as sites, but the edits here are about balance and feel, so you usually need more of them and you have to play between each one.

1. Theme: "replace the gems with fruit, warm palette, light background". One sentence changes the entire look.
2. Board size: "make the board seven by seven instead of eight". Size changes difficulty because it changes how many moves exist.
3. Number of gem types: "use five types instead of six". The fastest single lever for making the game easier or harder.
4. Move economy: "start with thirty moves, and give one extra move for every cascade of three or more".
5. Win condition: "instead of a score goal, require fifteen red gems collected". That is two or three edits rather than one, because the goal display and the end of run check both change with it.

Change one thing and then play. Match-3 is unusually sensitive to balance: a single parameter changes the feel of a whole run, and there is no way to judge it except by playing. Every rebuild opens in a sandbox before it reaches you, so a bad edit costs 3 credits and a minute rather than an evening staring at a black screen. The general rhythm of chat editing is in how to edit after generation.

What does reshaping Gem Match cost?

Cloning a game template is free: the page is copied into your project without a model call, so the first playable run costs nothing at all.

Free comes with 5 credits, enough for the clone and one edit, which is enough to see your own theme on the board. Pro is $25 a month with 200 credits, roughly 66 edits, and Business is $99 a month with 800 credits.

Budget more edits for a game than for a site. A landing page is finished when it reads well, but a game is finished when it feels right, and feel is found by trying. A retheme plus a balance pass is usually somewhere between eight and twenty edits, which still fits inside a single Pro month.

What should you not change without thinking?

Three pieces of this template carry more weight than their size suggests, and removing them tends to be regretted.

The attract mode is the first. A game that plays itself on the title screen converts far better than a screenshot, because a visitor understands the rules before deciding whether to press anything. If you restyle the title screen, keep it running underneath.

The hint is the second. It appears after a few idle seconds and points at a legal move, which quietly rescues the player who cannot see one. Removing it makes the game feel harder without making it more interesting.

The screen shake and the combo banners are the third. They exist because a match without feedback feels like nothing happened. You can retune them, and you should if they feel excessive on a phone, but stripping them out leaves a game that technically works and reads as flat. There is more on what a browser game can be in can AI build a game.

What are the honest limits here?

Zugo builds 2D games that run in a browser, and Gem Match is a good example of the ceiling rather than an exception to it.

There is no sound in the template. Music and effects can be added, but audio in a browser has its own rules, including the one where nothing plays until the player interacts with the page, so treat it as a small project rather than a one line edit.

Progress is not saved. The best score lives in memory for the session and disappears on reload, and there are no levels stored anywhere. Persistent progress, profiles or a leaderboard mean a database and sign in, which is a Supabase job on top of the game.

And there is no monetisation, no ad slots and no in app purchases. The template is a complete playable game, not a games platform. Anything with real player accounts and real money attached stops being a template edit and becomes a product with a roadmap, and that is where a builder stops replacing a development team.

Where to start

Open the gallery, switch to games, find Gem Match and play a run in the preview before cloning anything. Because previewing and cloning are both free, the only cost of trying all five game templates is your time.

Then work in this order: theme, board and gem count, move economy, win condition. Doing the visual pass first is tempting and usually wasteful, because a balance change afterwards can invalidate half of it.

The game is at zugo.dev, free credits included, and once you publish it you get a link you can send to a friend and watch them play, which is the only feedback that really tells you whether the balance is right.