Gradium builds low-latency speech recognition, voice generation and conversational-audio models for production voice agents. It was created by researchers behind Kyutai's real-time speech systems to turn generative-audio research into developer infrastructure.
See something inaccurate or outdated?
