OpenJarvis: Personal AI, On Personal Devices

Jon Saad-Falcon, Avanika Narayan, Robby Manihani, Tanvir Bhathal, Herumb Shandilya, Hakki Orhun Akengin, Gabriel Bo, Andrew Park, Matthew Hart, Caia Costello, Chuan Li, Christopher Ré, Azalia Mirhoseini
Preprint 2026

Materials

paper code blog demo

Abstract

OpenJarvis is an open-source framework for personal AI agents that runs entirely on-device. The system provides composable primitives for building on-device agents while treating efficiency metrics like energy and cost as primary evaluation targets alongside accuracy.

Coverage & Adoption

Selected industry and community references to OpenJarvis.

Industry & Analyst

  • Slingshots // Three: Our Third Grant Cohort
    Laude Institute · Jun. 2026
    read
  • OpenJarvis Review 2026: Stanford's Local-First AI Agent Framework
    ComputerTech · Mar. 2026
    read

Talks & Videos

  • Maximizing Intelligence per Watt (Podcast with Avanika Narayan & Jon Saad-Falcon)
    Foundation Capital · Jul. 2026
    watch
  • OpenJarvis: Personal AI, on Personal Devices
    AMD Advancing AI Day (YouTube) · Jul. 2026
    watch