The field of multi-agent collaboration and planning is moving towards more efficient and adaptable systems, leveraging large language models (LLMs) to enable intelligent collaboration and decision-making. Researchers are exploring novel frameworks and architectures that balance adaptability and efficiency, providing sufficient planning horizons while avoiding expensive full re-planning. Noteworthy papers include ELHPlan, which achieves comparable task success rates while consuming only 24% of the tokens required by state-of-the-art methods. Prompting Robot Teams with Natural Language presents a framework for prompting multi-robot teams with high-level tasks using natural language expressions, enabling decentralized and interactive real-time operation. AIPOM introduces a system supporting human-in-the-loop planning through conversational and graph-based interfaces, enhancing user control and trust in multi-agent workflows. TACOS presents a unified framework for high-level natural language control of multi-UAV systems through LLMs, integrating natural language interface, intelligent coordinator, and autonomous agent capabilities.