One-command deploy on any substrate
Run on bare metal, LXD, VMs, or Kubernetes with the same Juju workflow.
Deploy Hermes Agent on any cloud with Juju — from bare metal and LXD to VMs and Kubernetes — with a polished dashboard, multi-provider AI routing, and operational actions built in.
juju deploy hermes
Everything you need to run Hermes Agent as a managed, cloud-native service.
Run on bare metal, LXD, VMs, or Kubernetes with the same Juju workflow.
Anthropic, OpenAI, Google, GitHub Copilot, OpenRouter, DeepSeek, Groq, xAI, and more.
Configure resilient fallback behavior so Hermes keeps responding even when a primary model is unavailable.
Use the dashboard for day-to-day management and open a terminal-style chat experience when you need it.
The gateway and dashboard are supervised and started as managed services for reliable operation.
Recover state with Juju actions and keep the deployment fresh with update support.
Connect Hermes to Telegram, Discord, Slack, WhatsApp, Signal, LINE, Matrix, Mattermost, and more.
Expose the controls you need for status, dashboard URLs, backup, models, and pairing management.
A clean interface with professional motion, strong contrast, and mobile-friendly layout.
Deploy Hermes, then configure your primary provider and fallback chain.
juju deploy hermes
juju config hermes ai-provider=copilot ai-model=claude-sonnet-4 ai-api-key=ghu_xxx
juju config hermes ai-provider=anthropic ai-model=claude-sonnet-4 ai-api-key=sk-ant-xxx
juju config hermes ai-fallback-models="openai/gpt-4,google/gemini-2.5-pro"
Use the built-in Hermes CLI over Juju SSH for interactive chat.
juju ssh hermes/leader -- sudo -u ubuntu hermes chat --tui
juju ssh hermes/leader -- sudo -u ubuntu hermes chat
The web dashboard auto-starts on port 9119 and is reachable at the unit IP.
Operational hooks exposed through Juju actions.