Jump to:
Introduction
There are 2 main voice types in the AI Voice Generator: AI Voices and LLM Voices. AI Voices are created using deep learning and neural networks and are considered the app’s standard voices.
"LLM" is a next generation type of AI technology that uses Large Language Models in deep learning algorithms to create voices. The result is the highest quality voices you can find on the market today with realistic tones, emotions, and contextual understanding. LLM Voices are also multilingual, so you can use the same voice for 29 different languages.
LLM Voices include the Multilingual and Cloned categories in the AI Voice Generator’s voice settings. Learn more about Voice Cloning here.
Each paid user is given 6 million credits to use per month for TTS (text to speech) conversion. How many credits are used per character depends on if you are using an AI or LLM Voice.
TTS (Text-to-Speech) Credits and Limitations for Paid Users
Voice types in the AI Voice Generator are either AI or LLM. LLM Voices include Multilingual and Cloned Voices. Keep track of your monthly usage by going to Account > TTS Usage.
All paid plans include credits you can use to spend on either AI or LLM Voices. Credits are consumed every time a character is converted to audio. How many credits each character consumes depend on if you are using an AI or LLM Voice.
6,000,000 credits/user/month included
AI Voices = 1 credit/character
LLM Voices = 20 credits/character
If you run out of monthly characters, you can purchase additional Pay-As-You-Go credits to extend your usage.
Unused credits from the default 6 million monthly allotment do not carry over to the next month.
What counts as a character?
Everything that passes through the text-to-speech service counts as a character, including:
All letters and numbers
All spaces, punctuation, tabs, symbols, and special characters
All text included in SSML markup (Advanced Tunes tags)
Chinese, Japanese, and Korean language characters are counted as 2 characters
When does it reset?
Your credit limit resets month-to-month from the date of your subscription start date at 0:00 UTC. If you reach your monthly limit, you can purchase additional credits.
Changing your text, voice, speed, Pronunciation Editor settings, or Advanced Tunes settings requires that the text be re-converted to audio. When you modify the text or settings in any way, the characters in the edited text block will be counted again towards your usage limit.
Turbo Engine
LLM Voices now support Turbo Engine. When Turbo Engine is enabled, it prioritizes faster processing for text-to-speech conversion, delivering results more quickly. Although this may slightly reduce audio fidelity, it ensures efficient and timely output while also using fewer credits.
Toggle Turbo Engine
Open and login to the AI Voice Generator
Select Voices from the menu
Select Settings
Click to select turning Turbo Engine OFF or ON
Supported Languages for the AI Voice Generator
Languages supported by AI Voices:
English (US, UK, Australia, India), Arabic (Multi-region, Egypt, Saudi Arabian), Chinese (Simplified, Traditional, Taiwanese), Danish (Denmark), Dutch (Netherlands), Filipino (Philippines), Finnish (Finland), French (France, Canada), German (Germany), Hebrew (Israel), Hindi (India), Icelandic (Iceland), Indonesian (Indonesia), Italian (Italy), Japanese (Japan), Korean (Korea), Norwegian (Norway), Polish (Poland), Portuguese (Portugal, Brazil), Romanian (Romania), Russian (Russia), Spanish (Spain, Latin America, Mexico, US), Swedish (Sweden), Thai (Thailand), Turkish (Turkey), Ukrainian (Ukraine), Vietnamese (Vietnam), and Welsh (Wales)
Languages supported by Multilingual and Cloned Voices:
English, Arabic, Bulgarian, Chinese, Croatian, Czech, Danish, Dutch, Filipino, Finnish, French, German, Greek, Hindi, Indonesian, Italian, Japanese, Korean, Malay, Polish, Portuguese, Romanian, Russian, Turkish, Slovak, Spanish, Swedish, Tamil, and Ukrainian