agent swarm
A multi-agent setup in which many AI agents work on separate parts of a task in parallel and their results are combined.
An agent swarm is an approach in which a large task is split into smaller subtasks that separate AI agents handle at the same time. Each agent typically works in its own isolated context containing only the information its subtask needs, and the outputs are gathered and merged into a single result.
It is discussed as one form of multi-agent system, alongside a single agent working through tasks in sequence and hierarchical setups in which a supervisor agent delegates to sub-agents and summarizes their output. Key design choices include how results are merged and how work boundaries between agents are drawn.
This entry is based on AIPOST articles and widely known facts. If something is wrong, please send us a correction request.
Articles covering this entry
▲ Hierarchy and roles inside an agent swarm