# Deploy Hermes Agent with Easypanel

URL: /templates/hermes-agent

Hermes Agent is an autonomous AI agent by Nous Research that runs persistently on your server, remembers what it learns, executes shell commands, manages files, browses the web, schedules cron tasks, and integrates with Telegram, Discord, Slack, and WhatsApp. It supports 200+ models through OpenRouter and direct integrations with Anthropic, OpenAI, Google, and DeepSeek. The built-in web dashboard provides a browser-based UI for managing configurations, API keys, MCP servers, and chat sessions. All sessions, memory, skills, and configuration are stored in a persistent volume so the agent grows more capable over time.

## Installation notes

The Hermes Agent web dashboard is accessible at your assigned domain. Log in with the dashboard username and password you configured during deployment. The OpenAI-compatible gateway API runs on port 8642 and the dashboard backend runs on port 9119 — both are published for direct access. To get a model working, set up a model provider (OpenRouter, Anthropic, OpenAI, Google, or DeepSeek) from the dashboard or via the CLI. For messaging platforms like Telegram, Discord, Slack and WhatsApp, and especially setting up strict allowlists so only you (or your team) can message the agent — SSH into the service console and run hermes setup (or hermes gateway setup specifically for the messaging gateways).

## Benefits

### Web Dashboard Included

A built-in browser-based dashboard lets you manage API keys, MCP servers, messaging channels, memory, and chat sessions without touching the CLI. Protected by basic-auth and available at your domain immediately after deployment.

### Persistent Memory & Skills

The agent remembers your conversations, builds skills from experience, and retains context across sessions — all stored in a persistent volume on your own server.

### Multi-Provider Model Support

Works with 200+ models via OpenRouter plus direct integrations with Anthropic, OpenAI, Google, DeepSeek, and more. Switch models at any time with hermes model or through the dashboard.

### Messaging Gateway

Connect the agent to Telegram, Discord, Slack, or WhatsApp so it responds to messages, runs scheduled tasks, and delivers results to your phone — no open ports required.

## Features

### Browser-Based Admin Dashboard

Configure settings, manage credentials, browse the MCP catalog, and chat with the agent directly from a React-based web UI served at port 6432.

### OpenAI-Compatible API Server

Port 8642 exposes an OpenAI-compatible gateway so Claude Code, Open WebUI, Cursor, or any OpenAI SDK client can connect to Hermes as their LLM backend.

### Autonomous Tool Use

Built-in tools for shell execution, file operations, web search, and code running — the agent decides which tools to use and chains them to complete tasks.

### Cron Scheduling

Schedule recurring tasks with natural-language prompts and have results delivered to your messaging platform of choice.

### MCP Server Support

Connect any MCP-compatible server for additional tool integrations, or expose Hermes itself as an MCP server to Claude Code, Cursor, or other editors.

## Configuration

- `appServiceName` (App Service Name, required): hermes-agent

- `appServiceImage` (App Service Image, required): nousresearch/hermes-agent:v2026.6.5

- `dashboardUsername` (Dashboard Username, required): admin

- `dashboardPassword` (Dashboard Password): Set during installation — Basic-auth password for the web dashboard. Auto-generated if left blank — find it in the service environment variables after deployment.

## Services

- hermes-agent: app (nousresearch/hermes-agent:v2026.6.5)

## Resources

- [Website](https://hermes-agent.nousresearch.com/)

- [Documentation](https://hermes-agent.nousresearch.com/docs)

- [Github](https://github.com/NousResearch/hermes-agent)

- [Template source](https://github.com/easypanel-io/templates/tree/main/templates/hermes-agent)