New models: stt-rt-v6, stt-async-v6
Soniox · Product launch · November 3, 2026
Soniox introduced its v6 speech-to-text models, featuring agent text input and voice echo cancellation to improve accuracy and reduce false interruptions for AI voice agents. The update also enhances multilingual recognition and speaker diarization.
AntSpace’s summary of Soniox’s own announcement: read the original.