FreeUnsure how to say a name or word you just read? Select it and hear it instantly.

Add to Chrome
Developer API · Live

A free dictionary API that returns everything in one call

GET/v1/dictionary/:word

Definitions, usage examples, part of speech, synonyms, antonyms, IPA for US and UK, and a syllable breakdown with stress marks. One GET request, one small JSON response, one free API key by email.

Free in public testingOne key by emailOpen CORS116,800+ words
Request an API keyFull API reference
curl https://api.quickpronounce.site/v1/dictionary/wonderful
Why this one

More than definitions

Most free options give you a definition and stop. This one is built for apps that also need pronunciation and word relationships.

One request, not three

Definitions, examples, synonyms, antonyms, IPA, and syllables all come back together, so you're not calling a thesaurus service and a phonetics service on the side.

One free key, browser-callable

Email for a key, add one header, done. Open CORS means an extension or a single-page app can then hit it directly, with no OAuth and no backend proxy to stand up.

Pronunciation baked in

Every word carries IPA for US and UK and an anglicized syllable respelling with stress, not just a definition.

A predictable envelope

Every response is { success, data } or { success, message }. Easy to parse, easy to type, easy to hand to a language model as a tool.

Response

What comes back

A single word lookup returns this shape. Fields are always present unless noted.

Response fields
FieldTypeDescription
wordstringThe looked-up word, normalized to lowercase.
entries[]arrayOne object per part of speech. Each has partOfSpeech, definitions[], and examples[].
defaultPartOfSpeechstringThe most common part of speech for the word, handy when you only want to show one sense.
synonyms / antonymsstring[]Related words, ordered roughly by relevance. Present for about half the dataset.
phonetics.us / phonetics.ukstringIPA transcription for each accent, always populated.
syllables.us / syllables.ukarrayAnglicized respelling. Each item is { text, stress } where stress is 1 for the primary-stress syllable and 0 otherwise.
GET /v1/dictionary/wonderful
{
  "success": true,
  "data": {
    "word": "wonderful",
    "entries": [
      {
        "partOfSpeech": "Adjective",
        "definitions": [
          "Tending to excite wonder; surprising, extraordinary.",
          "Surprisingly excellent; very good or admirable."
        ],
        "examples": [
          "They served a wonderful six-course meal.",
          "What appears to be wonderful may turn out to be anything but."
        ]
      },
      {
        "partOfSpeech": "Adverb",
        "definitions": ["Exceedingly, to a great extent."],
        "examples": ["The project was wonderfully successful."]
      }
    ],
    "defaultPartOfSpeech": "Adjective",
    "synonyms": ["great", "nice", "top", "fine", "fabulous"],
    "antonyms": ["boring", "dull", "horrible", "common"],
    "phonetics": { "us": "/ˈwʌndɚfl/", "uk": "/ˈwʌndəfl/" },
    "syllables": {
      "us": [
        { "text": "wuhn", "stress": 1 },
        { "text": "derfl", "stress": 0 }
      ],
      "uk": [
        { "text": "wuhn", "stress": 1 },
        { "text": "duhfl", "stress": 0 }
      ]
    }
  }
}
Quick start

Call it from anywhere

No SDK. Add your X-API-Key header, then copy, paste, run. These snippets match the ones in the API repo's examples folder.

curl -H "X-API-Key: YOUR_KEY" \
  https://api.quickpronounce.site/v1/dictionary/wonderful
OpenAPI specLive playground
Failure modes

Errors and rate limits

Both failure modes use the same envelope as a success, so one code path handles them.

A request with no X-API-Key header, or an invalid key, returns 401. A word that isn't in the dataset returns 404 with { "success": false, "message": "Word not found." }. A route that doesn't exist returns the same shape. When you cross the per-key threshold you get 429 along with RateLimit-Limit, RateLimit-Remaining, RateLimit-Reset, and Retry-After headers, so a well-behaved client can wait exactly as long as it needs to.

Limits are 30 per minute per key, plus a per-key daily quota you can read at GET /v1/me. They're fair-use thresholds during public testing, not a hard global ceiling, and they're all environment-overridable if you run a self-hosted copy. If you need more headroom for a real workload, email [email protected].

Failure responses
# Missing or invalid key -> 401
$ curl -i https://api.quickpronounce.site/v1/dictionary/wonderful
HTTP/2 401
{ "success": false, "message": "Missing X-API-Key header." }

# Unknown word -> 404, same envelope
$ curl -i -H "X-API-Key: YOUR_KEY" https://api.quickpronounce.site/v1/dictionary/asdfghjkl
HTTP/2 404
{ "success": false, "message": "Word not found." }

# Too many requests -> 429, with standard headers
HTTP/2 429
RateLimit-Limit: 30
RateLimit-Remaining: 0
RateLimit-Reset: 41
Retry-After: 41
{ "success": false, "message": "Rate limit exceeded." }
In the wild

What people build with it

Anywhere you need lexical data without licensing a commercial dictionary.

Vocabulary and flashcard apps

Pull a definition, two example sentences, and an IPA string in a single request, then build a card. No stitching together three different services.

Browser extensions

Open CORS means a content script can call the endpoint directly on double-click or hover and render a definition popover, with no backend of your own.

Writing assistants and editors

Offer inline synonyms and antonyms as the user types, or surface a plain definition on hover for uncommon words.

Chatbots and AI agents

Give a language model a real dictionary tool so it stops guessing at definitions and pronunciations. The JSON envelope is small and predictable to parse.

Language-learning platforms

Attach a definition, usage examples, and a syllable breakdown to every word in a lesson without licensing a commercial dictionary.

Accessibility tooling

Read-aloud and simplification tools can fetch a plain-language definition and a phonetic guide for words a reader is likely to stumble on.

FAQ

Frequently Asked Questions

Straight answers on cost, limits, licensing, and how this compares to the alternatives.

Is the dictionary API really free?
Yes. It's free while the API is in public testing. You need an API key, which is granted on request with no billing attached, and it covers the full dictionary endpoint. Paid plans for higher request volumes are planned for later, once usage shows where the limits should sit. Free tier users get advance notice before any limit changes.
Do I need to sign up or get a key?
You need a key, but there's no signup form. Email [email protected] and one is sent back. Then pass it as an X-API-Key header on every request to the dictionary and pronunciation endpoints.
What are the rate limits?
The dictionary endpoint allows 30 requests per minute per key, plus a per-key daily limit you can check at GET /v1/me. These are fair-use thresholds while the API is in public testing, not a hard billing ceiling. RateLimit-* and Retry-After headers are sent on every response so your client can back off politely.
How many words are covered, and where does the data come from?
About 116,800 words and 178,600 dictionary entries. Definitions and examples are derived from Wiktionary (CC BY-SA 4.0) and rewritten, with pronunciation data drawing on CMUdict and WordNet. Coverage is 100 percent for IPA and syllables, 96 percent for at least one example, and about 52 percent for synonyms.
Can I call it straight from the browser?
Yes. CORS is open, so fetch() works from any origin with no proxy once you send the X-API-Key header. Several other free dictionary APIs block browser calls, so this is a real difference if you're building a client-side app or an extension.
How is this different from the Free Dictionary API or WordsAPI?
The Free Dictionary API (dictionaryapi.dev) is keyless but returns no synonyms or antonyms for many words and no anglicized syllable respelling. WordsAPI is richer but requires a key and a paid plan past a small free quota. QuickPronounce sits in between: a single free key by email, then IPA for US and UK, a syllable breakdown with stress marks, synonyms, antonyms, and optional synthesized audio from the pronunciation endpoint.
Can I cache responses?
Yes, and you should. The data changes rarely. Caching by lowercased word for hours or days is fine and keeps you well under the rate limits. There are no per-response cache restrictions.
Is commercial use allowed?
Yes, within fair-use limits. Because definitions and examples derive from Wiktionary under CC BY-SA 4.0, keep an attribution line in any product that reproduces them (the wording on the main API page is a good template). Bulk redistribution of the dataset is a separate licensed product, not something API access grants.
Is there an SDK?
Not yet, and you probably don't need one. It's a plain GET request that returns JSON. The OpenAPI spec at api.quickpronounce.site/openapi.yaml works with code generators and with AI coding tools if you want typed clients.
How stable is the response shape?
The /v1 envelope is stable. New fields may be added, but removing or renaming a field, or changing its type, is a breaking change that would ship under /v2 with a deprecation notice, never silently on /v1.

Keep Reading

Pronunciation APIAdd a synthesized MP3 clip to any word, with accent, gender, and speed controls.Phonetic / IPA APIIPA for US and UK plus an anglicized syllable respelling with stress marks.Full API referenceEvery endpoint, the live playground, rate limits, and the dataset breakdown.

Build on it today

The API is free while it's in public testing: email [email protected] for a key and you're set. The free tier works for prototypes, side projects, and small production apps. Paid plans for higher request volumes are planned as usage grows. The full enriched dataset is also available as a one-time licensed download, separate from API access.

Email for an API keySee all endpoints

Dictionary content is derived in part from Wiktionary, used under CC BY-SA 4.0, with definitions and examples rewritten and enriched by QuickPronounce. Pronunciation data draws on CMUdict and WordNet.