With super speed enabled, it painted this picture in only 14 seconds. At normal speed it QA-tests the app’s new dark theme, clicking through every tab to prove it works.
Hermes Desktop vs ServerEngine
Hermes Desktop gives you one self-improving agent in a window of its own, pointed at any model you like. ServerEngine gives you a workspace of agents on your own Windows machine, with the approvals, the scopes and the skills already built in.
Hermes Desktop
Feature by feature
| Hermes Desktop | ServerEngine | |
|---|---|---|
| Models | Any provider, including local models. | Claude Code, Codex CLI, Grok Build, OpenRouter or Ollama, per agent. |
| Your existing subscription | Signs in to Nous, Codex, Copilot or Anthropic, or takes a key. | Your Claude, ChatGPT or Grok plan, or an OpenRouter key. |
| Getting started | Installers that set up the agent and its runtime. | A signed Microsoft Store app that updates itself. |
| Agent Workspace | Chats in a sidebar, split panes and agent profiles. | Many agents side by side, each on its own model. |
| Talk or type | Live voice conversation, with a wake word. | Speak your prompt, and hear the answer when it is done. |
| Memory | Memory across sessions, and a graph you can edit. | One memory, shared by every model. |
| From your phone | Telegram, Discord, Slack, WhatsApp, Signal and more. | Slack and Discord per agent, WhatsApp for the main chat. |
| Agents on a schedule | Recurring jobs, set up in plain language. | Any prompt on a timetable, with its own model and skills. |
| Plans you can follow | Tools can ask before they run. | Plan mode, a live checklist, and approvals per skill. |
| Code editing | File and terminal tools, with a live projects view. | The fastest code editor for agents: built in pure C++, with batched calls so edits land in a fraction of the time. |
| Built-in skills | An open skills catalogue, skills it writes itself, and MCP. | Code, browser, computer use, images, GitHub, Stripe and more. |
| Work on your servers | SSH and Docker backends you set up yourself. | Your servers and a VPS too, on the Commander plan. |
| Keep it local | Runs on your hardware, with a local model if you like. | Runs on your PC. With Ollama, nothing leaves your network. |
Models
Hermes Desktop
Any provider, including local models.
ServerEngine
Claude Code, Codex CLI, Grok Build, OpenRouter or Ollama, per agent.
Your existing subscription
Hermes Desktop
Signs in to Nous, Codex, Copilot or Anthropic, or takes a key.
ServerEngine
Your Claude, ChatGPT or Grok plan, or an OpenRouter key.
Getting started
Hermes Desktop
Installers that set up the agent and its runtime.
ServerEngine
A signed Microsoft Store app that updates itself.
Agent Workspace
Hermes Desktop
Chats in a sidebar, split panes and agent profiles.
ServerEngine
Many agents side by side, each on its own model.
Talk or type
Hermes Desktop
Live voice conversation, with a wake word.
ServerEngine
Speak your prompt, and hear the answer when it is done.
Memory
Hermes Desktop
Memory across sessions, and a graph you can edit.
ServerEngine
One memory, shared by every model.
From your phone
Hermes Desktop
Telegram, Discord, Slack, WhatsApp, Signal and more.
ServerEngine
Slack and Discord per agent, WhatsApp for the main chat.
Agents on a schedule
Hermes Desktop
Recurring jobs, set up in plain language.
ServerEngine
Any prompt on a timetable, with its own model and skills.
Plans you can follow
Hermes Desktop
Tools can ask before they run.
ServerEngine
Plan mode, a live checklist, and approvals per skill.
Code editing
Hermes Desktop
File and terminal tools, with a live projects view.
ServerEngine
The fastest code editor for agents: built in pure C++, with batched calls so edits land in a fraction of the time.
Built-in skills
Hermes Desktop
An open skills catalogue, skills it writes itself, and MCP.
ServerEngine
Code, browser, computer use, images, GitHub, Stripe and more.
Work on your servers
Hermes Desktop
SSH and Docker backends you set up yourself.
ServerEngine
Your servers and a VPS too, on the Commander plan.
Keep it local
Hermes Desktop
Runs on your hardware, with a local model if you like.
ServerEngine
Runs on your PC. With Ollama, nothing leaves your network.
Any model
All superpowers
Run many agents side by side. Use your favorite models from Claude, Codex, Grok and OpenRouter while sharing the same memory and skills.
- ClaudeIntegrates with Claude Code CLI
- CodexIntegrates with Codex CLI
GrokIntegrates with Grok Build- OpenRouterIntegrates with OpenRouter API
- OllamaIntegrates with Ollama App
Memories grow on their own or whenever you ask. Your preferences, corrections and setup carry into every conversation, mapped to their work folder and shared with whichever model you run.
Send a message from your phone. Your agents keep working while you enjoy some free time.
Everything you need to build, test, secure, and ship your product. Enable what you need, disable what you don't, and select which skills require your approval.
Free
$0
Free access. No credit card.
The AI agent with core features
- S.E.N.T.R.I Agent
- Agentic Workspace (2 workspaces)
- Agents on a schedule (3 schedules)
- Multi-step task planning
- Core skills & approvals
- Coding agent integration
- Local speech-to-text
S.E.N.T.R.I
$15
per month, billed monthly
The AI agent & Workspace
- S.E.N.T.R.I Agent
- Agentic Workspace (unlimited)
- Agents on a schedule (unlimited)
- Multi-step task planning
- Core + DevOps skills & approvals
- Coding agent integration
- Local speech-to-text
- Voice responses
Commander
$30
per month, billed monthly
S.E.N.T.R.I + Servers
- Everything in S.E.N.T.R.I
- Zero-trustThe agent never gets a password. ServerEngine signs in for it from an encrypted store on your PC. agent access
- Agent Guardrails (ScopesA scope is a group of servers an agent may use. The rest stay hidden from it.)
- Remote Manager
- Admin Console
- RESTful API access