Hermes Desktop App
Beginner's Guide
Your complete guide to the native Hermes Agent desktop application. Install, configure, choose a model, and get your AI agent running in minutes.
Get the Free Guide
Compatible with macOS, Windows, and Linux. No credit card required.
What is the Hermes Desktop App?
The Hermes Desktop App is a native desktop application that wraps the full Hermes Agent in a purpose-built graphical interface. It shares the same config, API keys, sessions, skills, and memory as the CLI and gateway — it is not a separate product or a lightweight clone.
If you have used hermes in a terminal, everything you set up there is already in the desktop app. Anything you set up in the desktop app shows up in the CLI, TUI, web dashboard, and gateway. They all share state.
Whether you use the desktop app, CLI, TUI, or web dashboard, you are talking to the same Hermes Agent. Start a conversation in one and resume it in another. They are interchangeable frontends to the same engine.
All Hermes Interfaces
| Interface | Best For |
|---|---|
| Desktop App | Full native experience with chat, settings, file browser, and management panes |
| CLI (hermes) | Quick terminal chat, scripting, automation |
| TUI (hermes --tui) | Rich terminal UI with session browser and command palette |
| Web Dashboard | Browser-based admin panel with optional chat tab |
| Gateway | Mobile access via Telegram, Discord, and 24/7 autonomous operation |
Installation Three Ways
Three installation paths depending on your platform and whether you already use Hermes.
Option A: Download the Installer (macOS / Windows)
Visit hermes-agent.nousresearch.com
Click the “Download Desktop” button at the top of the page to get the latest installer.
Run the installer
On macOS, drag the app to Applications. On Windows, run the NSIS installer. First-launch setup is automatic.
Launch and follow onboarding
The first-run wizard walks you through adding a provider key and choosing a model in seconds.
Option B: CLI Install on Linux
curl -fsSL https://raw.githubusercontent.com/NousResearch/\ hermes-agent/main/scripts/install.sh | bash -s -- --include-desktop
Then launch with: hermes desktop
Option C: If You Already Have Hermes Installed
hermes desktop
Uses your existing config, API keys, sessions, and skills. Nothing to reconfigure.
Already a Hermes user? Option C is your fastest path. The desktop app inherits everything you have already set up in the CLI.
CLI Reference for Desktop
| Command | Description |
|---|---|
| hermes desktop | Launch the desktop app |
| hermes desktop --cwd PATH | Set initial project directory |
| hermes desktop --skip-build | Skip npm install, launch existing release |
| hermes desktop --build-only | Build but do not launch |
| hermes desktop --source | Launch via electron against dist |
The app checks for updates in the background and offers one-click updates. On macOS, reset a stuck microphone prompt with: tccutil reset Microphone com.nousresearch.hermes
Choosing a Model
The desktop app includes the same model provider configuration as the CLI, presented in a clean settings interface.
How to Add a Provider
- Open the desktop app and navigate to Settings > Providers
- Choose a provider: OpenRouter, OpenAI, Anthropic, Google Gemini, xAI, Ollama, or custom API
- Paste your API key (or use
hermes setup --portalfor one-click OAuth) - Select which model to use. The app tests the connection automatically.
Which Model Should You Start With?
| Your Needs | Recommended Model | Why |
|---|---|---|
| Best overall | Claude Sonnet 4 | Best balance of reasoning, tool use, and speed |
| Low cost, good quality | GPT-4o Mini | Fast, cheap, capable for most tasks |
| Open source / local | Qwen 2.5 32B | Run via Ollama, no API costs, fully private |
| Maximum performance | Claude Opus 4.8 | Top-tier reasoning for complex workflows |
| Quickest start | Nous Portal | One OAuth covers model + web search + image gen + TTS + browser |
Fastest path: run hermes setup --portal in the terminal. One OAuth flow gives you a model provider plus all four Tool Gateway tools.
First Things to Do
You have the app running and a model selected. Here is what to do next.
Send your first message
Type something simple: “What can you do?” Notice the streaming response, live tool activity, and structured tool-call summaries.
Explore the Settings panes
Check Providers & Keys, Toolsets (60+ tools), and MCP Servers. Everything is configurable from the UI.
Connect a messaging platform
Add Telegram or Discord under Settings > Messaging. Now your agent is accessible from your phone.
Try the file browser
Navigate to a project and ask Hermes to read a file or explain some code. The preview rail shows content side by side.
Schedule a cron job
Create a daily news briefing or email check. Cron runs even when the desktop app is closed.
Install skills from the catalog
Browse 50+ pre-built skills. Install one that matches a task you do regularly.
Save a fact about yourself
Tell Hermes: “Remember I work Eastern Time.” That fact persists across every session and interface.
Try voice mode
Click the microphone button. Speak your question. Hear Hermes respond.
Set up a second profile
Create separate profiles for work and personal. Skills, tools, and cron jobs stay isolated.
API-based providers cost real money per token. The desktop app shows no cost meter by default. Set usage alerts in your provider dashboard for intensive workflows.
Exploring the App
The desktop app is organized as a chat-first window with a left sidebar for navigation.
Chat
- Streaming responses with live tool activity and structured tool-call summaries
- Shared history — sessions started here resume in CLI/TUI/gateway
- Drag-and-drop files into the chat to attach them
- Preview rail — view web pages, files, and tool outputs side by side
File Browser
Explore and preview the working directory. Set the initial directory with --cwd.
Voice
Talk to Hermes and hear it respond. Same voice mode available in the CLI.
hands-freeSettings & Onboarding
Manage providers, models, tools, and credentials from a real UI.
no YAMLManagement Panes
Skills Browser
Browse, install, and manage skills from a visual catalog. Search by name or category.
50+ availableCron Manager
View, create, and manage scheduled jobs. See run history and output logs.
visual schedulerProfiles
Switch between Hermes profiles or create new ones. Each has isolated config and sessions.
multi-agentMessaging
Set up gateway channels — Telegram, Discord, Slack, WhatsApp, SMS, Teams.
8+ platformsPro Tips
Use the preview rail
When the agent generates code or renders HTML, the result appears in the right-hand preview rail. Keep chatting while you inspect.
Resume sessions across interfaces
Start in the desktop app. Continue from Telegram. Finish in the CLI. All share session history.
Set up cron early
Cron jobs execute on the backend regardless of whether the desktop app is open. Schedule an email check or daily briefing.
Drag files, don't type paths
Drag files into the chat area from your file system or the file browser. No path typing needed.
Build skills from what works
Save effective approaches as skills via the Skills pane. Skills compound in value over time.
Troubleshooting: Boot logs in ~/.hermes/logs/desktop.log. Tail with hermes logs gui -f. Reset with rm ~/.hermes/hermes-agent/.hermes-bootstrap-complete.
Getting Help
| Resource | Best For |
|---|---|
| Official Documentation | Every feature, config option, and reference |
| Skill Catalog | 50+ pre-built skills to install |
| GitHub Repository | Source code, issues, release notes |
| Discord Community | Community help, tips, and discussion (link on Hermes website) |
The Hermes community on Discord is active and helpful. Search the docs first, then ask in the community. The project is open source and rapidly evolving — new features land weekly.
About Viable Edge
Viable Edge provides brand management and marketing automation for early-stage SaaS companies. We help you build authority, automate content operations, and turn your expertise into a growth engine.
Ready to go deeper?
Viable Edge builds on Hermes Agent to create fully managed marketing and operations automation. Get a complete AI-powered content engine without the infrastructure complexity.
Visit Viable Edgeviableedge.com — Build your AI operations stack