Introduction
OM1 is A modular AI runtime for agents and robots with capabilities including movement and speech.
What is OpenMind (OM1)
OM1 allows AI agents to be configured and then deployed in both the digital and physical world. You can create one AI agent and run it in the cloud but also on physical robot hardware such as Quadrupeds, and, soon, TurtleBot 3 and Humanoids.
For example, an AI agent built on OM1 can ingest data from multiple sources (the web, X/Twitter, cameras, and LIDAR) and can then Tweet and explore your house, shake your hand, or talk to you. In another example, with OM1, you can talk with OpenAI’s gpt-4o and literally shake hands with it.
Capabilities of OM1
Simple, modular architecture
Easy-to-understand, independent components that work together seamlessly.
All python
Independent modules that are easy maintain, and extend.
Easy to add new data inputs
Seamlessly integrate new data without major changes to the existing structure.
Easy to support new hardware
via plugins for API endpoints and specific robot hardware
Can be connected to:
ROS2, Zenoh, and CycloneDDS
Includes a simple web-based debug display
to watch the system work (WebSim at http://localhost:8000)
Preconfigured endpoints
for Voice-to-Speech, OpenAI’s gpt-4o, DeepSeek, and multiple VLMs
Was this page helpful?