Ant Group and Tsinghua open-source Realtime-Venus, a full-duplex multimodal agent alternative to OpenAI's Realtime API
aigclink · x · 2026-10-16
Ant Group's Venus team and Tsinghua University have open-sourced Realtime-Venus (Apache-2.0), a real-time multimodal agent system positioned as an open-source alternative to OpenAI's Realtime API.
- It can continuously see, listen, and speak while delegating long-running tasks to background execution — aimed at smart glasses, in-car assistants, and similar always-on scenarios.
- Its dual-loop architecture is built on two 9B models plus a Harness:
- Realtime-Venus-Omni: continuous audio/video perception;
- Realtime-Venus-Audio: native full-duplex voice;
- Realtime-Venus-Harness: asynchronously runs reasoning, search, and tool calls, then returns results to the ongoing conversation for spoken delivery.
- Front-end stays responsive in real time while heavy work runs in parallel, so users never interrupt a conversation just to wait for task completion.
GitHub: github.com/inclusionAI/Realtime-Venus
More from coding & agent
- OpenAI discloses GPT-5.6 Sol derived wallet keys and spent unauthorized funds — terryyuezhuo · 2026-10-16
- Vibe-coded sonar FPS fits a full game in 16 KB with shader ray-casting — zTrojan · 2026-10-16
- OpenAI chart: coding agents now log 3.14x human researcher workdays; Anthropic merges 30k changes/day — haider1 · 2026-10-16
- Supabase CTO on how AI agents will reshape software buying and open source — David Ondrej · 2026-10-16
- Dev builds MCP server that lets AI agents outsource phone calls to real humans — Many_Benefit_3393 · 2026-10-16
- A vector DB is not LLM memory: long thread maps write/read paths, failure modes and evals — techNmak · 2026-10-16