“With rich voices, limitless customizations, and high-quality audio,” Meta notes this will “unlock deeper storytelling, more believable conversations, and ultimately deeper immersion.”
A separate update from Meta’s Reality Labs team provides more technical detail. ElevenLabs is now available as a pre-integrated text-to-speech (TTS) and speech-to-text (STT) provider within Meta’s new AI Inference Building Blocks — a system designed to let developers plug external AI models directly into Quest applications without building custom integrations.
ElevenLabs brings a sizable multilingual library to the collaboration, with 11,000+ voices across 70+ languages, spanning text-to-speech, dubbing, and AI music models. The company says its models “adapt to every tone, accent, and culture” — positioning the integration as a way for Meta to “make diverse audio a core layer of their AI experiences.”
With the ElevenLabs partnership, Meta is making voice AI a built-in component of both its consumer products and its developer ecosystem.
First Reactions
Early reactions to the partnership spanned enthusiasm, creator concerns, and strategic interpretation.
Developers welcomed the move, noting that high-quality voice generation has long been a bottleneck for building immersive VR experiences. Others highlighted that AI dubbing removes one of the last barriers to global content distribution, allowing creators to publish once and reach everywhere.
Some described the integration as part of a broader shift in which AI audio becomes product infrastructure, not a standalone feature — enabling new “creative and communicative patterns.”
Slator Market Recap 2025
The Slator 2025 Language Industry Recap covers key developments in 2025 across language AI, localization, AI dubbing, voice AI, and more.
However, others found this development “slightly alarming” and expressed caution about platforms automatically modifying creators’ content. Concerns centered on loss of control, cultural nuance, and tone — especially when AI translations are applied without creator review. Even technically accurate AI translations, they noted, may not capture humor, sentiment, or regional speech patterns.
Some offered a strategic interpretation of what they called a “smart move.” They suggested the partnership may reflect Meta’s need for a high-quality voice solution today while its internal voice models mature. Under this “rent-to-own” view, Meta deploys ElevenLabs immediately to stay competitive with TikTok and YouTube’s rapidly advancing dubbing features, while continuing to develop its own LLM-based voice systems in parallel. In the meantime, Elevenlabs becomes embedded into the world’s largest short-form distribution engine, gaining both revenue and brand recognition.