AISiliconANGLEabout 2 hours ago
PolyAI launches real-time voice conversation model
1 min read
PolyAI released Dialog-RSN-1, a voice AI model that processes audio directly instead of using transcripts. The model achieves 280-500ms latency, undercutting OpenAI's GPT-realtime-2.1 at 860-1900ms. A separate text-to-speech module gives companies control over voice customization and compute costs.
Level
Hype check
Tap to vote and see what everyone thinks.
Summary by ByteBrief
