Below you’ll find answers to some of the most common questions about generating voiceovers. These FAQs cover everything from supported file formats and recommended resolutions to storage, privacy, and content management, helping you get the best results and keep your creations secure.
What are the file requirements for Custom Voices?
The file limit is 20MB, minimum of 10 seconds, maximum of 5 minutes, and the supported files types are mp3, WAV, or m4a. For more information read here.
What languages are voiceovers available in?
The languages available to use when generating a voiceover are model-dependent. Look at the details of different models here.
What languages are Custom Voices available in?
Custom voices can generate audio in all supported languages, except Swedish.
Can team users access the custom voices of other team members?
Voices uploaded by one team member are visible and usable by the entire team. For teams and businesses.
How many characters can my prompt be?
You can type up to 5000 characters for each voiceover generation.
How do I know I can switch to a different model if the language I want isn’t supported?
Unsupported languages in the selected AI Model (but available in another AI Model) appear disabled with a switch icon. Clicking the Switch icon will change the model to one that supports the selected language.
What file format is the downloaded voiceover?
Voiceovers can be downloaded as an MP3.
Before the Toolkit, the voiceover feature had V1 and V2. What are the equivalents to them in the Toolkit?
In the Toolkit, V1 is the equivalent to Cartesia Sonic 2, and V2 is the equivalent to MiniMax 02 HD.
Can I use voiceover tags to direct performance?
Eleven v3 can use audio tags to direct performance. Write any text in [brackets] before the text the tag is meant to direct. Tags can be any text and written in any language.
For example:
"[Fast] Good morning!"
"[Calm] Good morning!"
How can I get a British or Australian accent in a voiceover when using Eleven Multilingual v2, Eleven v3, or MiniMax 02 HD?
The accent selection feature is only available when using the Cartesia model.
If you’re using a different model, you can still achieve a British or Australian accent by choosing one of our voices that already have those accents built in. In this case, no additional accent settings or customization are required.
- Esteem — British
- Wit — British
- Charmed — British
- Professor — British
- Precision — British
- Grounded — British
- Posh — British
- Quest — British
- League — British
- Santa — British
- Views — British
- Tattle — British
- Unfiltered — Australian
- Guidance — Australian
Still need help? Head back to the Help Center.