Tsinghua University's THUNLP lab, in collaboration with OpenBMB and others, has released PilotDeck 2.0, an open-source operating system for AI agents. This system addresses key challenges in agent performance, including managing multiple projects, providing transparent and editable memory, enabling background task execution, and optimizing model usage through intelligent routing. PilotDeck 2.0 aims to reduce costs by up to 83% for complex tasks by intelligently assigning simpler sub-tasks to lower-cost models while reserving high-end models for critical decision-making. AI
IMPACT This release could significantly lower the operational costs of AI agents and improve their usability for complex, multi-project workflows.
RANK_REASON This is a software release for AI agents, not a frontier model release or core research paper.
- Claude Code
- Claude Opus 4.5
- Claude Sonnet 4.5
- Claude Sonnet 4.6
- MiniMax M2.7
- OpenBMB
- PilotDeck 2.0
- Qwen3.6 35B-A3B
- THUNLP
- Tsinghua University
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
