Early Learning Garden · Free & open source

Count. Listen. Build. Hop.

A jar of ladybugs, a robot with a word to say, a garden of letters and a bunny finding its way. Four small games for a child and a grown-up to play together.

No account to play · Local saves · No AI subscription

Early Learning Garden paper-and-felt illustration: ladybugs, a robot making SUN, letter flowers and a bunny on a numbered path.
Illustrated with ChatGPT. All four games are included.
01 · How Many?

A quick look. A little count.

Spot the ladybugs before the cover comes down. Start with simple patterns, then try scattered groups and clusters up to six. Look again whenever you need.

Count ladybugs →

02 · Robot Words

Slide the sounds together.

Move along a word one sound at a time, then choose its picture. Twelve short words, three stages, and a slider that works with touch, mouse or keyboard.

Build a little word →

03 · Letter Garden

Find the sound.

Listen to a modelled sound and choose its letter. Eighteen letters arrive in five waves, with more practice for the ones that need it.

Visit the letter garden →

04 · Bunny Hops

Count on from here.

Help the bunny move one space at a time along a numbered path. Later rounds start farther along, so counting does not always begin at one.

Help the bunny →

Pull up a chair

A grown-up supplies the sounds.

The games include prompts and guidance, with no prerecorded voice pack. For a letter sound, say the sound shown in Sound and grown-up help, then press Sound modelled, continue. Say the sound, rather than the letter’s name.

You can add your own MP3 recordings from a local folder. An optional local device voice can read ordinary instructions and whole words; it does not model phonemes. Recordings stay on your device and must be selected again after reloading or changing games.

These are original, AI-assisted personal learning games. The stages are design choices, not a validated curriculum or a claim about school readiness.

From a link to a game of your own

No computer skills required.

Choose a game is the easiest route, especially on a phone. On a computer, you can keep the whole collection for offline play.

  1. Download the offline collection below. Save the ZIP in Downloads or somewhere easy to find.
  2. Extract the ZIP. On a Mac, double-click it. On Windows, right-click and choose Extract All. Keep the five HTML files together.
  3. Open Early-Learning-Garden.html. Double-click it, then choose a picture. If it opens as text, choose Open with and select your browser.
  4. Play together. Start as Player 1 or add a nickname in Learners and backups. A grown-up models letter sounds, or adds recordings.
Download all four games →

You do not need a GitHub account, a terminal, an API key or a subscription.

Keep your progress

A backup belongs to you.

Each game saves its own learners and progress in this browser. Use Export this learner’s progress in each game before clearing browser data, moving files or changing browsers. A backup contains the nickname and progress in readable text, so keep it private. Import creates a separate learner.

If browser storage is unavailable, the game says Temporary session. You can still play and export before leaving. Competing tabs preserve the losing attempt for recovery.

The games have no analytics or default cloud connection. Optional cloud saves are included for people hosting their own copy, with adult email sign-in. Choose which learner to attach or restore; once attached and connected, later saves upload automatically. The offline files keep cloud disabled.

Built for me. Yours to adapt.

Make it your own.

I built this for me and my personal use. Make it your own, and feel free to improve mine. Hopefully it gives you a useful starting point, or at the very least some ideas. Cheers!

The complete games, content banks, recording tools, design notes, tests and an AI-assistant adaptation skill are on GitHub. This page uses a pinned release of that same public engine.

The MIT license lets you use, change, share or sell your version, keeping the notice. Dependencies keep their own notices. Playing needs no AI service or tokens.