Edster releases open-source local AI agent with swarm mode — lets developers run multiple AI workers on their own computers without cloud dependency

An open-source project called Edster (GitHub: unrealumanga/Nedster) introduced a local AI agent framework with 'swarm mode'—the ability to spin up multiple AI workers that coordinate tasks together, plus a web interface for managing them, all running on your own hardware instead of rented cloud servers.

Unlike cloud-based AI tools (OpenAI, Claude API) that charge per request and send your data to external servers, Edster runs entirely on local machines, so you pay no per-use fees and keep all data private—particularly useful for companies handling sensitive information or those operating in regions with poor internet.

Software developers and small teams can now build AI-powered automation (chatbots, data processing, workflow automation) without monthly API bills or vendor lock-in; the swarm mode means one person's code can orchestrate dozens of parallel AI tasks, useful for processing large datasets or handling multiple customer requests simultaneously.