Key Features:

Audiobox is a collection of 6 AI driven demos that allow users to create voices and sound effects using voice recordings and text prompts, in addition to a timeline feature where these generations can be arranged and remixed. 

Pros:

  • UI is simple and easy to navigate. 
  • Voice and sound effect descriptions can be incredibly specific, accounting for the speakers age, accent, and environment they are in. 
  • Generated results can be instantly downloaded and rated to improve the AI’s accuracy in the future. 
  • Audiobox Maker allows users to create and customize fully AI generated stories with multiple sound layers. 
  • Generations include audio watermarking that traces sounds back to their sampled recordings to promote responsible AI.

Cons:

  • Odd pronunciation and pacing in generated voices are easy giveaways that voices are AI produced. 
  • Magic Eraser tool can be finicky and cannot use previously recorded files. 
  • Small limit of 30 words per text prompt is unideal for longer scripts. 
  • Only 2 results per search and lacking in advanced settings.

Pricing:

  • Free and accessible to all.

Use Cases for Enterprise

Internal training/experimentation, supplemental tool for Media Industry and marketing.

Detailed Review

Meta’s evolution of their AI model Voicebox, Audiobox is an AI driven audio generation demo suite, meant to further research in the world of AI generated speech and sound. Audiobox comes with easy-to-use tools for users to create AI generated voices from a customizable text prompt, from a sample of the user’s voice, or a hybrid of the two. What is most impressive about Audiobox is its ability to implement prompts that dictate the sound quality, background sounds, and even the emotional tone and accent of the speaker. Audiobox also has a standalone sound effects generation demo as well as an audio story creator known as Audiobox Maker, where generated voices and sound effects can be layered and arranged on a timeline.  

While the prompt length limitations and limited result selections of these tools make it harder to recommend as a comprehensive tool to professional media companies, Audiobox is a perfect starting point for those new to AI to experiment with what is possible with AI generated audio and serves as an example of what an AI driven sound design workflow could look like.