Accent Translation (AT5.2) and Language Translation Are Now Available
Two new Human ↔ Human models are live on the Sanas Developer Platform:AT5.2(Accent Translation): Modifies accents in real time so every speaker is instantly understood, while preserving what makes their voice their own.- LT Documentation (Language Translation): Translates conversations across languages in real time, preserving each speaker’s tone and intent with no interpreter delay.
Get a Developer Account, Faster
Developer accounts can now be created through self-serve signup, no sales contact required.
Two New Models Replace AGENTIC_SE
AGENTIC_SE is deprecated and replaced by two specialized models:AGENTIC_VI_G_NC(Default): Removes environmental background noise and distant human voices from the primary speaker’s audio streamAGENTIC_ST_NC: Removes environmental background noise while preserving distant human speech
Migration
Updateaudio_params.modelName from "AGENTIC_SE" to either new model. No other code changes needed.