OpenMind/OM1 is a Go-based modular AI runtime for multimodal agents and robotics, with ROS2/Zenoh integration, pre-configured endpoints, and Prometheus/Grafana observability. Latest releases include new MCP tools, cron scheduling, and Unitree Go2 support.
Collecting history — the radar snapshots this repo daily. The trend line appears after 3 days of data (1 so far).
What it is
OpenMind's OM1 is a modular AI runtime written in Go that enables developers to create and deploy multimodal AI agents across digital environments and physical robots, with hardware support via plugins for ROS2, Zenoh, and CycloneDDS.
How it works
OM1 provides a core agent pipeline in Go, with pre-configured endpoints for text-to-speech, multiple LLMs from various providers, and multiple Visual Language Models. It includes a pre-configured Prometheus and Grafana stack to monitor metrics such as LLM and ASR latencies.
Getting started
Quick Start (5 Minutes)
- Install system dependencies.
- Download the binary or clone the repository.
- Add your OpenMind API key.
- Launch OM1 and verify it is running.
1. Install System Dependencies
macOS:
brew install portaudio ffmpeg
Linux:
sudo apt-get update
sudo apt-get install -y portaudio19-dev ffmpeg
2. Download or Build
Option A: Download Pre-built Binary (Recommended)
No Go installation required. Download the latest release for your platform from the Releases page. Current pre-built binaries are available for:
linux-amd64linux-arm64darwin-arm64darwin-amd64
After downloading and extracting the archive, open a terminal in the extracted folder and run:
chmod +x om1
On macOS, if the binary is blocked by Gatekeeper, run:
xattr -d com.apple.quarantine om1 2>/dev/null || true
The release archive already includes
config/,knowledge_base/, andlibzenohcfiles, so you do not need to clone the repository for runtime assets.
Option B: Build from Source
Requires Go 1.23.0+ and make.
git clone https://github.com/OpenMind/OM1.git
cd OM1
make deps
make build
3. Configure API Key
Get your API key from OpenMind Portal.
Set via shell profile (recommended):
export OM_API_KEY="<your_api_key>"
If you built from source from this repository, you can also use a project-local .env file:
cp .env.example .env
# Then set OM_API_KEY=<your_api_key> in .env
4. Launch the Agent
If using pre-built binary: On macOS:
export DYLD_LIBRARY_PATH="$PWD:$DYLD_LIBRARY_PATH"
./om1 -config ./config/conversation.json5
On Linux:
export LD_LIBRARY_PATH="$PWD:$LD_LIBRARY_PATH"
./om1 -config ./config/conversation.json5
If you run into permission issues on macOS, adjust privacy settings accordingly. If built from source:
CONFIG=conversation make run
Or for development with debug logging:
CONFIG=conversation make dev
5. Monitor with Grafana (Optional)
docker-compose up -d grafana prometheus
Navigate to http://localhost:3000 with default login admin/admin. The OM1 Latency Monitoring dashboard is provisioned.
Release notes (latest 5)
- nightly Development Build (2026-06-04): added latest release notes for om1 and om1-video-processor; Increase default max_turn_count to 5.
- v1.0.2-beta.2 (2026-04-29): Add Unitree Go2 patrol background plugin; Add conversation mode and person follow functionality.
- v1.0.2-beta.1 (2026-04-14): Introduce MCP (Model Context Protocol) tools with reference configs; Add cron job scheduling; Add G1 arm action support.
- v1.0.1 (2026-03-15): Unifies single and multi modes and improves stability of mode switching.
- v1.0.1-beta.3 (2026-03-12): Single mode migrated to multi-mode; multiple feature fixes.
Traction
Stars: 2895. Forks: 992. Open issues: 92.
License
MIT
Repository metadata
Language: Go Created: 2025-01-08 Last push: 2026-08-04 Topics: llm, multiagent, robotics, ros2, zenoh