# Stanford's DeLM cuts multi-agent task costs 50% - without a central orchestrator

> **Platform:** [SPIDITS AI](https://spidits.com/) — Real-Time AI News & Market Intelligence  
> **Published:** 2026-06-16T17:47:53.000Z  
> **Category:** ACQUISITION  
> **Impact Score:** 250/100  
> **Primary Source:** [VentureBeat](https://venturebeat.com/orchestration/stanfords-delm-cuts-multi-agent-task-costs-50-without-a-central-orchestrator)  
> **Canonical Citation:** [https://spidits.com/timeline/stanfords-delm-cuts-multi-agent-task-costs-50-without-a-central-orchestrator](https://spidits.com/timeline/stanfords-delm-cuts-multi-agent-task-costs-50-without-a-central-orchestrator)

## Executive Summary
One of the assumptions behind today's AI frameworks is that agents require a "boss" at the center; this orchestrator runs the show, routes requests, and makes...

## Referenced Coverage & Sources
- **[VentureBeat](https://venturebeat.com/orchestration/stanfords-delm-cuts-multi-agent-task-costs-50-without-a-central-orchestrator)**: Stanford's DeLM cuts multi-agent task costs 50% - without a central orchestrator — _One of the assumptions behind today's AI frameworks is that agents require a "boss" at the center; this orchestrator runs the show, routes requests, and makes..._

---
*Synthesized by SPIDITS AI Market Intelligence Desk. Track live AI news, model releases, and funding: [https://spidits.com](https://spidits.com)*