Pronunciation Dictionary - Get Every Name Right

A pronunciation dictionary tells the AI voice exactly how to say your brand names, people's names, acronyms and technical terms, and every generation after that follows it. No phonetic alphabet is needed: you write how the word should sound, and even in a script tens of thousands of characters long, the same name is read the same way every time.

Mispronounce the brand once and the whole spot is unusable

The most embarrassing AI voiceover failure isn't the wrong tone — it's getting your product name wrong. nginx read letter by letter, SaaS spelled out, a founder's surname mangled. In a long script those errors repeat dozens of times.

A pronunciation dictionary fixes exactly that: tell the model how a word should be read, and every generation obeys.

Why AI voices mispronounce words

A speech model has to guess each word's pronunciation from its spelling and context. These are the words it gets wrong most often:

• Acronyms: some are spelled out, some are said as a word; SaaS is usually said "sass"
• Technical terms: words like nginx are rare in everyday text, so the model has little to go on but the spelling
• Brand names and surnames: an invented brand name or an uncommon surname such as Ng has no obvious reading
• Words with two readings: live sounds different as an adjective and as a verb

Rather than rewriting the script or regenerating until it comes out right, add one rule to a pronunciation dictionary.

Real A/B

The same English line, before and after applying four rules:

Without the dictionary

ボイス: 甜美少女英語4秒

With the dictionary

ボイス: 甜美少女英語3秒

Text: Our SaaS runs on nginx. Ship a GIF with VoiceSmiths.

Rules:
  SaaS       → sass
  nginx      → engine ex
  GIF        → jif
  VoiceSmiths → Voice Smiths

Writing pronunciation dictionary rules

A dictionary is a written-form → spoken-form replacement table. No phonetic alphabet required — just spell it how you want it said:

A few habits that save time:

1. Respell by sound: break the word into parts that read naturally, like engine ex or Voice Smiths
2. Test on one short line first: confirm the reading, then use it on the full script
3. Mind words with two readings: before adding a rule for a word like live, check that the rest of the script uses the same reading
4. Split dictionaries by purpose: one for brand terms, one for people, combined per project as needed

How to use it

1. Studio → Pronunciation: name the dictionary, add rules, create
2. Back on the voiceover tab, dictionaries appear as chips — tick the ones to apply
3. Stack several at once (e.g. "Brand terms" + "People"), applied in order

Where it matters most

• Long-form audiobooks: a proper noun appearing a hundred times gets fixed once
• Brand spots and product videos: one wrong product name and the whole spot needs redoing
• Technical courses: software names and acronyms stay consistent from the first lesson to the last
• Audio dramas: character and place names run through every episode, so listeners never get confused

Versus emotion tags

• Emotion tags like [whispers] control how it's acted
• The dictionary controls how it's pronounced

Use both together — they don't conflict, and neither leaks into subtitles.

FAQ

Do I need to know phonetics to use the pronunciation dictionary? No. A dictionary is a written-form to spoken-form replacement table, so you just spell each word the way you want it said, for example SaaS as sass.

Can I use several dictionaries at once? Yes. Stack as many as you need, such as "Brand terms" plus "People", and they are applied in order.

Will the respelled words show up in my subtitles? No. The dictionary only changes how a word is pronounced; the respelling never leaks into the subtitles.

Does it work together with emotion tags? Yes. Emotion tags control how a line is acted and the dictionary controls how words are pronounced, so they never conflict.

Create your first pronunciation dictionary

Put the three to five words that get mispronounced most often into a dictionary, and hear the difference on your next generation.

Create my first dictionary →