AgentBeats Phase 1 Deadline & New Benchmarks
This week, we wrap up the first phase of the AgentBeats coding competition. We also introduce a new set of benchmarks for evaluating multi-agent collaboration in dynamic environments...
Prize Updates & Custom Tracks for OpenDev
Detailed breakdown of the prize pool for the new Custom Track. Also, we discuss the implications of multi-agent collaboration in cybersecurity environments and what it means for safety...
2025: A Year of Agentic Innovation
Looking back at our major milestones: The launch of ChatGLM-4, the first Agentic AI MOOC, and over 50 top-tier conference papers accepted. What's next for 2026?