Audiobox vs Voicebox by Meta
Side-by-side comparison · Updated April 2026
| Description | Audiobox is Meta’s innovative foundation research model for audio generation. It enables users to generate voices and sound effects with ease by using voice inputs and natural language text prompts. Audiobox includes specialized models such as Audiobox Speech and Audiobox Sound, which are built upon the self-supervised Audiobox SSL model. It provides a platform for users to create custom audio for various applications. Interactive demos, Audiobox Maker, and research information are available to explore its capabilities further. | Meta AI researchers have unveiled Voicebox, a cutting-edge generative AI model for speech that sets new standards in the field. Voicebox leverages a novel approach called Flow Matching to learn from raw audio and transcriptions, enabling it to modify any part of a given audio sample. It has outperformed existing models like VALL-E and YourTTS in terms of intelligibility, audio similarity, and processing speed. Voicebox has been trained on 50,000 hours of public domain audiobooks in multiple languages and can perform diverse tasks such as cross-lingual style transfer, noise removal, and content editing. Despite its capabilities, the model or code is not publicly accessible due to potential misuse, though Meta has shared audio samples and research papers detailing its functionalities. |
| Category | Audio Editing | Voice Modulation |
| Rating | No reviews | No reviews |
| Pricing | N/A | N/A |
| Starting Price | N/A | N/A |
| Use Cases |
|
|
| Tags | voicessound effectsvoice inputsnatural language text promptsaudio generation | generative AI modelspeechFlow Matchingraw audiointelligibility |
| Features | ||
| Generate voices and sound effects | ||
| Voice input and text prompt integration | ||
| Audiobox Speech for speech generation | ||
| Audiobox Sound for sound effects generation | ||
| Built on Audiobox SSL self-supervised model | ||
| Interactive demos available | ||
| Audiobox Maker for audio stories | ||
| Fairness and safety guardrails | ||
| Watermarked outputs for security | ||
| English language support | ||
| Generative AI for speech | ||
| Flow Matching technique | ||
| Zero-shot text-to-speech | ||
| Cross-lingual style transfer | ||
| Noise removal | ||
| Content editing | ||
| Multiple language support | ||
| State-of-the-art performance | ||
| 50,000 hours of training data | ||
| Not publicly available due to ethical considerations | ||
| View Audiobox | View Voicebox by Meta | |