Polya AI launches Dialog-RSN-1: realtime voice model with sub-300ms median first-token latency

matthen2 · x · 2026-10-20

• Dialog-RSN-1 is now generally available, with API keys open to anyone, plus an OpenAI Realtime-compatible API.
• Claimed numbers: median first-token latency under 300ms (about the length of a blink), with the highest quality, audio-understanding and safety scores among real-time-capable models, and efficiency to scale to every conversation.
• The model is audio-native: turn-taking, speech recognition, function calling and the response happen in a single pass, with separately controllable voice output.
• Three onboarding paths: the Agent Development Kit (ADK), AWS SageMaker, or the public API.

Original post →

More from Models

• PretrainGym pits frontier agents on pretraining data curation: Kimi K3 wins at 1.042 — xeophon · 2026-10-20
• SlideOn model card details SAE pretraining and Matryoshka sparse training recipe — tomaarsen · 2026-10-20
• LightOn ships 16.3M training pairs and full code alongside SlideOn models — tomaarsen · 2026-10-20
• SlideOn lands in sentence-transformers with one-line SparseEncoder loading — tomaarsen · 2026-10-20
• mSlideOn handles cross-lingual retrieval with raw dot-product scoring — tomaarsen · 2026-10-20
• mSlideOn tops MTEB Code at 74.78 nDCG, beating dense and late-interaction models — tomaarsen · 2026-10-20

Models channel →