OneMeta AI debuts Verbum: real-time voice translation for up to 50 people at once

Virtual worlds love to promise a borderless future where anyone on Earth can share the exact same space. The dream holds up nicely until people actually try talking to each other.

Services like Google Translate or Skype handle one-on-one chats reasonably well. But throw dozens of people from different countries into the same voice room, and communication instantly collapses. Each person speaks their own language, and traditional translation tools hit a wall.

At CES, OneMeta AI debuted a service called Verbum, specifically designed to tackle this multi-person mess.

50 voices, 82 languages

Verbum can handle real-time translation for up to 50 people on a single call, with everyone speaking their native tongue. The system currently supports 82 languages and 40 dialects.

Crucially, it is not just dumping text onto a monitor. Verbum handles speech too.

OneMeta spokesperson David Politis painted the picture: you could put 50 people on a Zoom call, and every single person could speak their native language. If someone speaks Japanese, another participant hears it in Italian, English, or Russian, while simultaneously seeing the translated text on their screen.

Politis pointed out that while English remains the corporate default, non-native speakers often struggle to convey nuance. Misunderstandings naturally follow, and Verbum wants to wipe out that friction entirely.

Testing the delay

In a hands-on demo at CES, the setup proved surprisingly capable.

During a headset call with a woman in Central America, English remarks were translated into Spanish for her, while her Spanish replies were converted back to English. Despite a slight pause, the conversation flowed naturally without awkward disruption.

Words were transcribed on screen within a second of being spoken. Around one second after that, the AI voice kicked in. It sounded remarkably human, easily matching or beating the familiar voiceover styles found on platforms like TikTok.

While OneMeta is targeting multinational business meetings first, the implications stretch directly into high-intensity gaming and metaverse hangouts, from chaotic MMORPG squad chats to global esports watch parties.

Why it matters

Most commercial translation tools treat language like a two-way street, connecting person A to person B. Verbum treats it like a multi-lane roundabout, coordinating real-time mesh translation for up to 50 people with roughly a two-second total turnaround for synthetic voice.

For remote international teams and global gaming communities, that leap in group dynamics is genuinely practical. It removes the tax of stumbling through broken second languages without requiring expensive live interpreters.

Still, OneMeta faces an obvious giant in the room. Real-time translation is a software feature that mega-platforms like Microsoft and Google could easily incorporate into Teams or Meet using their own deep infrastructure.

To survive beyond the initial demo buzz, OneMeta will need to prove its latency holds up under real network stress and figure out how to embed seamlessly where users already hang out.