VoxenAI Introduces Multilingual Voice Style Transfer, Raises $22M Series A

AI · 5 min read

VoxenAI released a multilingual voice style transfer API capable of rendering a speaker’s voice in another language while preserving intonation and style, coupled with a $22 million Series A.

The company’s model separates content, prosody, and timbre to generate natural-sounding multilingual speech without requiring large paired datasets. VoxenAI positions the product for dubbing, international podcasts, and accessible voice interfaces.

Privacy features include explicit consent flows and ephemeral embedding storage; partners can opt for on-prem deployments for sensitive content. Early clients include media houses and e-learning platforms trialing subtitled and dubbed archives.

Funding will scale GPU inference clusters, add language support for low-resource tongues, and expand SDKs for NLE and localization tools.