Levanto Labs launches Sage2: extract + decide models process live chats in ~200ms
AccBalanced · x · 2026-11-03
Levanto Labs unveiled Sage2, a new primitive for real-time conversation automation built on two models: an Extract stage that turns each incoming message into structured fields in 150ms with exact source spans (claimed 166x cheaper than Opus 5.5, beating Cloudflare's Clairic), and Decide, billed as the first decision model with streaming input, which scores options as text arrives and returns typed answers with probabilities in 200ms end-to-end. Unlike Jev-style fixed Q&A, it never re-reads the full transcript, enabling fully reactive automation on live chats where details emerge gradually. 20M free tokens for developers; also available via nearai and agent payment rails.
Related event: Levanto Labs Launches Sage2, a Low-Cost Real-Time Decision Model(2 posts)→
More from coding & agent
- LetterBox adds granular API key permissions, agents can create Drafts over MCP — rchardkovacs · 2026-11-03
- Tsinghua study: continual harness self-evolution causes severe catastrophic forgetting in coding agents — Shangziqi Zhao · 2026-11-03
- Vantage: open-source Windows app lets AI agents drive your real desktop, 80-step task in ~8 min — AKGAMING1234 · 2026-11-03
- Reward hacking risks when agents optimize inference, new article warns — AccBalanced · 2026-11-03
- Build a Media Agent workshop: CPU-based agent calling GPU tools on demand — Al_Grigor · 2026-11-03
- Factory rebuilt its Droid coding agent to run inside customers' own clouds, plans to open source the compute layer — AI Engineer · 2026-11-03