Hermes Studio

Web dashboard for Hermes Agent — multi-platform AI chat, session management, scheduled jobs, usage analytics

Last verified:

Visit Hermes Studio

What is Hermes Studio?

Hermes Studio is a self-hosted AI chat dashboard and desktop app for Hermes Agent, the autonomous AI agent system from Nous Research. It brings chat, runtime management, automation, files, coding agents, and release-ready desktop distribution into one visual workspace, turning the Hermes Agent CLI into a full command center for AI operations.

Key features include streaming Agent Chat with sessions, Markdown rendering, tool traces, file uploads/downloads, model badges, and context usage tracking; a built-in browser terminal via node-pty and xterm.js; a workspace file browser with directory tree, inline previews, and git detection; a skills marketplace to browse/install/uninstall Hermes skills; persistent memory across sessions via user profile and agent notes; self-hosted cron/scheduled jobs that fire offline and deliver to Telegram/Discord/Slack/email; multi-provider model support (OpenAI, Anthropic, Google, DeepSeek, OpenRouter, Ollama, LM Studio); session management with pin/archive/tags/projects; voice input via Web Speech API; 7 built-in themes; optional password authentication; Docker/Docker Compose support; and mobile responsive layout with Tailscale support.

Hermes Studio is designed for developers, AI researchers, and Hermes Agent users who want a visual control plane instead of terminal-only interaction. It suits anyone managing multiple AI agent profiles, scheduling recurring agent jobs, monitoring token usage and costs, connecting Hermes to messaging platforms, or working with local models via Ollama/LM Studio/vLLM.

Hermes Studio pricing

Pricing model: Freemium

Fully free and open-source under the MIT License for self-hosted use. Hermes Web UI can be installed and used for personal, educational, and research purposes under BSL-1.1. Commercial use requires a separate license until the project changes to Apache-2.0 on May 10, 2029. The desktop app is free to download with builds for macOS, Windows, and Linux. Docker images are pre-built and multi-arch (amd64 + arm64) published to GHCR. No subscription or paid tier for the self-hosted version - you pay only for your own AI model API costs.

Hermes Studio pros

  • Fully free and open-source under MIT License
  • Self-hosted on your own hardware with full data control
  • Full three-panel web UI layout (sidebar, chat, workspace)
  • Streaming SSE responses with auto-reconnect
  • Persistent memory across sessions
  • Self-improving skills system with 90+ built-in capabilities
  • Multi-provider model support (OpenAI, Anthropic, Google, DeepSeek, OpenRouter, Ollama, LM Studio)
  • Session management with pin, archive, tags, and projects
  • Workspace file browser with inline preview and git detection
  • Self-hosted cron/scheduled jobs with multi-channel delivery
  • Voice input via Web Speech API
  • 7 built-in themes including OLED
  • Optional password authentication
  • Docker and Docker Compose support
  • Mobile responsive layout with Tailscale support
  • Built-in browser terminal via node-pty and xterm.js
  • Skills marketplace with install/uninstall/update
  • Token analytics by model, platform, and time range
  • CLI session bridge from hermes-agent SQLite store
  • Tool call cards with expand/collapse and JSON viewer
  • Mermaid diagram rendering
  • Thinking/reasoning display for Claude and o3
  • Approval card for dangerous shell commands
  • Dark and light theme toggle

Hermes Studio cons

  • Requires self-hosting setup on your own hardware
  • Still needs working Hermes Agent runtime installed
  • Many configuration variables can be complex
  • Not a hosted chatbot service
  • Desktop setups need separate Hermes Agent installation
  • Default login (admin/123456) must be changed immediately
  • Chat starts may fail if runtime not discovered
  • Model providers often require separate accounts
  • Commercial use requires separate license until Apache-2.0 in 2029

Frequently asked questions about Hermes Studio

Is Hermes Studio free?

Yes, Hermes Studio (Hermes Web UI) is fully free and open-source under the MIT License for self-hosted use. You can install and use it for personal, educational, and research purposes without any subscription fees. The only cost is your own AI model API expenses from providers like OpenAI, Anthropic, or Google.

Does Hermes Studio require Hermes Agent to be installed?

Yes, Hermes Studio is a web UI/dashboard for Hermes Agent and requires a working Hermes Agent runtime. The desktop and npm setups need Hermes Agent installed separately (source checkout, package install, or compatible Python environment). The Docker Compose setup can run a single-container deployment with integrated Hermes Agent.

Can Hermes Studio run local models?

Yes, Hermes Studio supports local models through OpenAI-compatible endpoints. It works with Ollama, LM Studio, and vLLM for local model hosting. You need to add a provider from a preset or custom OpenAI-compatible endpoint in model management, and the UI can discover models from auth.json and provider /v1/models endpoints.

What AI model providers does Hermes Studio support?

Hermes Studio supports OpenAI, Anthropic, Google, DeepSeek, Nous Portal, OpenRouter, MiniMax, Z.AI, Ollama, and LM Studio via a dynamic model dropdown. It works with any OpenAI-compatible backend and supports credential pools, OAuth logins, and provider model discovery.

How do I install Hermes Studio?

You can install via three methods: (1) Desktop app - download the installer from GitHub Releases for macOS/Windows/Linux, (2) npm - run 'npm install -g hermes-web-ui' then 'hermes-web-ui start' and open http://localhost:8648, or (3) Docker - run 'WEBUI_IMAGE=ekkoye8888/hermes-web-ui docker compose up -d' and open http://localhost:6060. Change the default login (admin/123456) immediately after first login.

Does Hermes Studio have persistent memory?

Yes, Hermes retains context across sessions via user profile, agent notes, and a self-improving skills system. Memory entries can be viewed/edited in the Memory panel, and you can configure memory providers. This means you don't need to re-explain yourself every session.

Can I schedule automated jobs with Hermes Studio?

Yes, Hermes Studio includes self-hosted cron/scheduled jobs that fire while you're offline. You can create, edit, pause, resume, delete, and manually trigger cron jobs with preset schedules and next-run time display. Results can be delivered to Telegram, Discord, Slack, Signal, email, and more.

What authentication options does Hermes Studio have?

Hermes Studio offers optional password authentication via the HERMES_WEBUI_PASSWORD environment variable, using signed HMAC HTTP-only cookies. It also supports token auth. The dashboard includes security features like CSRF tokens on 21 endpoints, SRI-pinned CDN resources, dynamic CORS, WebSocket origin verification, XSS escaping, and rate limiting.

Does Hermes Studio work on mobile devices?

Yes, Hermes Studio has a mobile responsive layout with hamburger sidebar, touch-friendly 44px targets, and full-height chat on phones. You can access it via Tailscale for a native app-like experience from your phone.

What's the difference between Hermes Studio Desktop and Hermes Web UI?

Hermes Studio Desktop is the desktop app wrapper that bundles the Web UI runtime and uses the native Hermes data location. Windows stores Hermes data under %LOCALAPPDATA%\hermes, macOS/Linux under ~/.hermes. The Web UI wrapper stores its own state under ~/.hermes-web-ui unless HERMES_WEB_UI_HOME changes that path. Both provide the same features, but Desktop has lower setup friction.

Categories

Use cases

Browse all AI tools on NeedAnAI