Skip to content

词光 · Word Atlas — Illustrated IELTS vocabulary with 7,298 words, UK/US audio and FSRS review. Open-source app + CC BY 4.0 dataset.

6.9 KB
Raw Blame

词光 · Word Atlas — an illustrated world of words

Word Atlas · 词光

See the word. Hear it. Make it yours.

7,298 IELTS words · Original illustrations · UK & US audio · Spaced repetition

Try Word Atlas ↗  /  48-second film  /  Get the dataset  /  简体中文

Code: MIT Dataset: CC BY 4.0


The Word Atlas iOS app beta is open and free: api.billswise.cn/app

Follow @yangleiocx3

Buy me a coffee

This project is open source and free, and I have put a lot of my own money and time into it. If it helps you, you can buy me a coffee.

WeChat Pay QR code
WeChat Pay
Alipay QR code
Alipay

Word Atlas is an illustrated IELTS vocabulary app for Chinese-speaking learners. Connect words with scenes and pronunciation, practise active recall, and return to them with FSRS spaced repetition. No account required. Learning progress stays on your device.

Version scope: This repository contains the open-source foundation of Word Atlas and its dataset. The hosted website evolves separately and may have different features. The code and dataset here remain available under their respective MIT and CC BY 4.0 licenses.

Preview · Learning loop · Quick start · Open dataset · Contributing

Preview

A vocabulary atlas you can listen to and practise with

Open-source edition: desktop vocabulary atlas
Daily study plan on mobile Illustrated word card with pronunciation and examples
A plan for todayA scene for every word

Actual screenshots of the repository edition. The header is a conceptual brand illustration.

Watch your vocabulary take shape

Vocabulary galaxy showing learning progress across topics

Explore 19 topic groups, rotate and zoom the galaxy, and select a star to open its word card. Brightness reflects learning status. Built with WebGL2, with a fallback for unsupported devices.

Watch the 48-second introduction

Watch the Word Atlas introduction

Play the full film with sound

Learning loop

Step What you do
See Explore an illustration, bilingual definitions, two original examples, collocations and usage notes.
Hear Switch between UK and US pronunciation and listen to example sentences. A subset includes audio for the second example.
Recall Practise meaning-to-word, illustrated cloze, listening recognition and spelling.
Review Let FSRS schedule reviews based on your history. Only the first answer counts toward accuracy.
Keep Track progress in the galaxy, practise typing separately, and export or restore your local progress.

Local first. Visited entries and media can be cached for offline use. Export a progress backup before moving to another device.

Listen to a sample: abandon — UK · US · Example.

Quick start

Requires Node.js 20+ and npm. This repository includes the full image and audio collection, so the initial download can take time.

git clone --depth 1 https://github.com/designfuchen/word-atlas.git
cd word-atlas
npm ci
npm run dev

Open http://localhost:5188.

npm test             # Unit tests
npm run check:atlas  # Dataset and media validation
npm run build        # Production build in dist/

For a subpath deployment: ATLAS_BASE=/words/ npm run build. See the deployment guide.

Stack: React 19 · TypeScript · Vite · ts-fsrs · WebGL2

Open dataset

Use the app as it is, or build your own learning tools, Anki decks and research projects.

Resource Repository edition Download / documentation
Entries 7,298, CEFR A1–C2, 19 topics JSON · CSV
Examples 14,596, everyday and IELTS contexts, with Chinese translations Schema
Illustrations 7,298, 512 px WebP plus thumbnails Images
Recordings 22,905, UK/US words and example sentences Audio
import json
from pathlib import Path

words = json.loads(Path("data/word-atlas.json").read_text(encoding="utf-8"))
b2_environment = [w for w in words if w["cefr"] == "B2" and w["topic"] == "environment"]

Provenance and limitations

Text and illustrations were generated for this project using AI; speech was synthesized with ElevenLabs. Automated structure and consistency checks have been performed, but the entire collection has not been manually reviewed. Definitions, examples, illustrations or pronunciation may contain errors. This is not an official IELTS list and does not promise score improvements.

Generation pipeline · Learning model · Scripts

Contributing

Please include the word ID, reproduction steps and expected result when reporting a problem. For vocabulary corrections, reliable dictionary references are welcome.

License

Scope License
Application code MIT
Vocabulary text, illustrations and word recordings CC BY 4.0; credit “Word Atlas contributors(词光)”
Dependencies and fonts Their respective licenses

Made by @designfuchen. Follow the build on X · @yangleiocx3.

If Word Atlas helps you, consider starring the repository or sharing it with another learner.