Telnyx just added Rime Arcana V3 to their voice AI platform, bringing native code switching between 10 languages (English, Spanish, French, German, Arabic, Japanese, Portuguese, Hindi, Hebrew, Tamil) to their Voice API and AI Agents.
The interesting part: this works as a single utterance without breaking voice identity. So when a customer switches from "Hello, I need help with mi cuenta" or "Can you help me avec mon problème?" mid sentence, you get the same voice maintaining prosody across completely different phonetic structures.
Most multilingual TTS solutions handle this by routing to different voice models per language (your agent sounds like different people) or adding latency from language detection. Arcana V3 processes mixed language input as one continuous stream.
This is part of Telnyx's hub approach. Instead of being locked into one TTS provider's strengths and weaknesses, you get access to multiple frontier models through the same API. Arcana V3 for multilingual, other models for different specializations.
Particularly useful for voice AI serving bilingual communities where code switching is the norm, not the exception. Available now across their Voice API, AI Agents, and standalone TTS.
The interesting part: this works as a single utterance without breaking voice identity. So when a customer switches from "Hello, I need help with mi cuenta" or "Can you help me avec mon problème?" mid sentence, you get the same voice maintaining prosody across completely different phonetic structures.
Most multilingual TTS solutions handle this by routing to different voice models per language (your agent sounds like different people) or adding latency from language detection. Arcana V3 processes mixed language input as one continuous stream.
This is part of Telnyx's hub approach. Instead of being locked into one TTS provider's strengths and weaknesses, you get access to multiple frontier models through the same API. Arcana V3 for multilingual, other models for different specializations.
Particularly useful for voice AI serving bilingual communities where code switching is the norm, not the exception. Available now across their Voice API, AI Agents, and standalone TTS.