Skip to main content

Generating AI Voiceovers

Artlist’s AI Voiceover tool lets you create high-quality, professional voiceovers in multiple languages and accents. Whether creating a tutorial, promotion or podcast, customize delivery with filters, effects, and different voices to match your content. Use Text to Voice to turn prompts into lifelike audio, Voice to Voice to transform your recordings with AI models, or even clone a voice, to use when generating a voiceover. This article explains how to use each feature and shares tips for the most natural results.

Stay up to date with our latest news on the blog.

How to select a voice

Selecting the right voice model ensures your narration or dialogue fits the tone and personality of your content.

  1. In the sidebar on the homepage, click AI Voiceover.
  2. Below the prompt-area, find the Voice Catalog.
  3. To filter voice models by gender, click Gender.
  4. To filter voice models by Video Category, click Video Category.
  • Health & Wellness
  • Commercials
  • Social Explainers
  • Tutorials
  • Trailers
  • Characters
  • Documentaries
  1. Hover over the voice model of your choice and click Select to apply the voice. You can always preview the style before selecting by clicking play on the preview.

💡 Pro Tip: Click the featured assets icon in the bottom-right corner to check out the assets used in the voice model preview.

How to use Text to Voice

Text to Voice lets you instantly generate high-quality AI voiceovers from written prompts. With a range of filters, voiceover effects, and emotions, you can easily shape the language, tone, and style of any project to create natural-sounding recordings.

Once you have selected the voice you want to use:

  1. Use the language dropdown in the prompt-area, to select your language.
  2. In the prompt-area, type the text you want to generate.
  3. For generations in English, select the accent dropdown and choose from American, Australian, British, and Indian accents.
  4. Select the Speed dropdown to choose from a range between 0.8x and 1.2x.
  5. Select the Emotion dropdown to choose which emotion you’d like the voiceover to convey (if any).
  6. Use the Effects dropdown, to alter the voice. Once you've selected an effect, you will be prompted to adjust the Effect strength.

Note: Adjusting the effect strength (%) will only impact the final voiceover, not the preview of the effect.

  1. Click Generate.

Note: If you began using Artlist before the introduction of our Version 2 voice model, you'll see the option to switch back to the original version in order to use your previous voice models for ongoing projects. To do this, click the Switch to V1 Model above the text box. (Only available for accounts created before the update.) 

💡 Pro Tip: Version 2 offers greater stability, improved sound quality, and enhanced clarity, with broader support for emotions and languages to help you create more natural, expressive voiceovers.

How to use Voice to Voice

Voice to Voice lets you upload your recording and instantly transform its tone and intonation using a variety of voice models, styles, and emotions—while preserving your original delivery.

  1. In the sidebar on the homepage, click AI Voiceover.
  2. On the AI Voiceover catalog, click Voice to Voice.
  3. Upload your audio file -drag and drop, or select the file from your computer.
  4. Select the Voice Model and Effect.
  5. Click Generate.

Note: Your credit balance is shown beneath the text input box

💡 Pro Tip: To get the best results, recorded audio should follow these guidelines:

  • Format: The format of your recorded file should be sent as mp3, WAV or OGG.
  • File size: The maximum file upload size is 30MB.
  • File length: The maximum file length is 5 minutes.

How to clone a voice 

Voice cloning allows you to create a custom AI-generated voice in the Voice catalog. With this feature, generate voiceovers in your voice or a cloned voice of your choice (with permission), ensuring consistent voices across projects while saving time. 

  1. Below the prompt-area, find the Voice Catalog.
  2. Click Clone a voice.
  3. Upload a file. Either select a file from your device or drag and drop one in.
  4. Name your voice.
  5. Add a description and more details (optional).
  6. Click Clone. The number of credits used will be visible.

💡Pro Tip: Toggle remove background noise to make your voice as clear as possible. 

Pro Tip: To get the best results, recorded audio should follow these guidelines:

  • Format: The format of your recorded file should be uploaded as mp3, WAV, or m4a.
  • File size: The maximum file upload size is 20MB.
  • File length: The minimum file length is 10 seconds. The maximum file length is 5 minutes.

Note: All users must confirm they have the necessary permission to use the uploaded voice via the check box. 

Note: Deleting a cloned voice does not delete the voiceovers generated with it. These remain accessible in history, but can't be used for future generations. 

How to manage your voiceovers

To locate and manage your generated voiceovers

  1. In the sidebar on the homepage, click AI Voiceover.
  2. Click My Voiceovers.
  3. Select a voiceover to:
    1. Download
    2. Get AI Powered Suggestions based on your prompt. Discover music, footage, and SFX to use in your project. (Text to Voice only)
    3. Add to your favorites
    4. Rename, delete, or add to an Artboard using the more icon

Tips to generate better voiceovers

Creating a high-quality voiceover starts with the right setup and approach. This section outlines key tips that help improve clarity, tone, and overall performance when generating a voiceover. Understanding these tips makes it easier to get consistent, professional-sounding results.

  • Text input format
    • Use sentences of short phrases
    • Group phrases with brackets [ ] for natural rhythm. Example: [Try the new Voiceover tool.] [It’s easy to use.]
  • Pronunciation
    • Match the language setting to the text
    • Ensure correct spelling
    • Use phonetic spelling for tricky words, names, or trademarks
    • Write acronyms as spoken: U → “you”, X → “ex”, A → “ae”,I → “eye”
  • Emphasis
    • Use punctuation to shape tone
    • Use two question marks ?? for questions
    • Avoid quotation marks for emphasis (only use them for actual quotes)
  • Pauses using a native accent:
    • Use commas or dashes - for short pauses
    • <#1.0#> (the number is in seconds)
    • Avoid repeating punctuation (e.g., ,, or --)
  • Pauses using a non native accent:
    • For longer pauses, insert: <break time="1s" />

Note: Using the "reset" button in the accent dropdown will reveal what that voice's native accent is: (insert screenshot)

  • Numbers and Symbols
    • Break up long numbers with commas or dashes
    • For dates, use: 01/26/2025 or January 26th, 2025
    • To read each individual character, for example in a long numerical string or alphanumeric code, type <spell> before and </spell> after the text. Example: <spell> 4454AB </spell>
    • Write symbols as words: Example: # → “hashtag”
  • URLs & Emails
    • Write how it should sound. Example: artlist.io → “Artlist dot eye oh”
    • Add a space before question marks after URLs. Example: artlist.io ?
  • Break your script into smaller sections
    • Generating shorter segments of your script can help you fine-tune pacing for each part, making the overall voiceover feel more controlled and natural.
    • Add breaks/pauses between sentences in your editing software as well

 

Commonly asked questions

What are the file requirements for Voice to Voice? 

The maximum file upload size is 30MB. The maximum file length is 5 minutes. The format of your recorded file should be uploaded as mp3, WAV or OGG. For more information, see Voice to Voice.

What are the file requirements for voice cloning?

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 see How to clone a voice.

What language is Voice to Voice available in?

Voice to Voice is available in English only.

What languages are available for Text to Voice?

Text to voice AI voiceover is available in English, French, Finnish, Japanese, German, Portuguese, Spanish, Korean, Polish, Italian, Dutch, Swedish, Turkish, Arabic, Russian, Greek, Romanian, Cantonese, Czech, Hindi, Mandarin, Ukrainian, Vietnamese and Swedish.

What languages is Voice cloning available in?

Cloned voices can generate audio in all supported languages, except Swedish.

How many characters can my script be?

You can type up to 2000 characters for each voiceover generation.

Can team users access the cloned voices of other team members?

For businesses and team accounts, voices uploaded by one team member are visible and usable by the entire team. 

Still need help? Head back to the Help Center.