wiktionary-sound

wiktionary-sound

Retrieve pronunciation audio files from Wiktionary and Wikimedia Commons.

wiktionary-sound searches Wiktionary entries, finds available pronunciation recordings, resolves their Wikimedia URLs, extracts metadata (language, speaker, accent), and selects the best pronunciation automatically.

Useful for:

• Language learning applications
• Dictionary apps
• Flashcards
• Pronunciation tools
• React / Next.js applications
• Educational software

Features

• Supports Wiktionary languages
• Finds real pronunciation recordings
• Extracts speaker information when available
• Detects accents from filenames
• Automatically selects the best pronunciation
• Built-in caching to reduce repeated requests
• Works with Node.js and modern bundlers
• Includes TypeScript declarations

---

Installation

Using npm:

npm install wiktionary-sound

Using pnpm:

pnpm add wiktionary-sound

Using yarn:

yarn add wiktionary-sound

---

Quick Start

JavaScript

import { getSound } from "wiktionary-sound";

const result = await getSound("bonjour", "fr");

console.log(result.best);

Example output:

{
  "title": "Fichier:Fr-bonjour.ogg",
  "url": "https://upload.wikimedia.org/...",
  "extension": "ogg",
  "language": "fr",
  "speaker": null,
  "accent": null
}

The best result is the recommended pronunciation file.

---

API Reference

getSound(word, language, options)

Searches Wiktionary for pronunciation files.

Parameters

Parameter	Type	Default	Description
word	string	required	Word to search
language	string	"en"	Wiktionary language edition
options	object	{}	Search options

Example:

const result = await getSound("chaussure", "fr");

---

Options

Language

The second argument controls the Wiktionary language edition.

Examples:

await getSound("hello", "en");

await getSound("bonjour", "fr");

await getSound("hola", "es");

---

Selection preferences

You can control pronunciation selection:

const result = await getSound("hello", "en", {
  preferLanguage: "en",
  preferAccent: "us",
});

Available preferences:

Option	Description
preferLanguage	Prefer recordings from a language code
preferAccent	Prefer a specific accent
preferNative	Prefer native pronunciation recordings

---

Response Format

getSound() returns:

{
  (word, sourceLanguage, files, best);
}

word

The searched word.

Example:

"bonjour";

---

sourceLanguage

The Wiktionary language used.

Example:

"fr";

---

files

All available pronunciation recordings.

Example:

[
  {
    "title": "Fichier:Fr-bonjour.ogg",
    "url": "https://upload.wikimedia.org/...",
    "extension": "ogg",
    "language": "fr",
    "speaker": null,
    "accent": null
  }
]

---

best

The automatically selected pronunciation.

Example:

{
  "title": "Fichier:Fr-chaussure.ogg",
  "url": "https://upload.wikimedia.org/...",
  "extension": "ogg",
  "language": "fr",
  "speaker": null,
  "accent": null
}

---

Selecting Pronunciations

The package uses a scoring system to choose the best audio file.

Higher priority:

1. Exact requested language match
2. Native language recordings
3. Official Wiktionary pronunciation files
4. Preferred accent match
5. Supported audio formats

Example:

For:

getSound("hello", "en");

Possible results:

En-us-hello.ogg
En-uk-hello.ogg
LL-Q1860-English recordings

The selector chooses the highest scoring result.

---

TypeScript

The package includes TypeScript definitions.

Example:

import { getSound, SoundResult } from "wiktionary-sound";

const result: SoundResult = await getSound("bonjour", "fr");

console.log(result.best.url);

---

React Example

The package can be used inside React applications.

Example:

import { useEffect, useState } from "react";
import { getSound } from "wiktionary-sound";

export default function Pronunciation() {
  const [audio, setAudio] = useState(null);

  useEffect(() => {
    getSound("bonjour", "fr").then((result) => {
      setAudio(result.best.url);
    });
  }, []);

  return audio && <audio controls src={audio} />;
}

---

Data Source

This package uses:

• Wiktionary API
• Wikimedia Commons audio files

All pronunciation recordings are provided by Wikimedia contributors.

---

Limitations

• Not every word has pronunciation audio.
• Some languages have fewer recordings.
• Audio quality depends on available Wikimedia files.
• Results depend on Wiktionary content.

---

License

MIT License

Copyright 2026 saugatproductions