Run Hermes Agent
on your own box —
in one click.
Hermes is the AI agent that grows with you — persistent memory, self-written skills, every chat app. Pick a plan. Hermes arrives on your own box, root access included, ready to chat.
no curl ｜ no setup wizard ｜ no server to babysit
Why deploy Hermes here
Hermes is meant to live on a server you own. You get exactly that — a dedicated box with Hermes already running. No curl commands, no systemd fiddling, no 2am "why did it go down" pages.
One click, zero install
No curl | bash, no setup wizard. Pick a plan — Hermes arrives on your dedicated box, configured, connected, and live.
It's still your box
Full root SSH. Hermes can run terminal commands, install packages, and harden itself — on real hardware, not a toy sandbox.
Dedicated IPv4 + IPv6
Static addresses that are yours alone. Whitelist them, peer with them, point your tools at an agent that always answers on the same IP.
Subdomain + TLS
Land instantly on hermes.you.agentdeploy.cloud with HTTPS handled — or map your own domain in a click.
Always on, always current
Runs as a managed service that survives reboots, with auto-updates and daily snapshots. The cron scheduler keeps firing while you sleep.
Memory that sticks
Hermes' ~/.hermes/ lives on a durable volume. Restarts, updates, and migrations never wipe what it has learned about you.
The hard way vs. the one click
Do it yourself
- Rent and configure a Linux server
- Run the install script, pray the deps resolve
- Set up the gateway as a systemd service
- Wire DNS, certificates, firewall rules
- Babysit uptime, updates, and backups forever
Deploy with agentdeploy
- Pick a plan
- Toggle the chat apps you want
- Click Deploy
- Get root, IPs, and your subdomain back
- Updates & backups handled, forever
Meet Hermes Agent
Not a chatbot. Not a copilot. A persistent agent that learns your projects, writes its own skills, and reaches you wherever you are.
Persistent memory
Remembers preferences, projects and context across every session.
Writes its own skills
Solves a hard problem once, saves a reusable SKILL.md forever.
Every chat app
Telegram, Discord, Slack, WhatsApp, Signal & CLI from one gateway.
Scheduled automations
Cron built in — daily briefings, nightly backups, weekly audits.
Parallel sub-agents
Spawn isolated agents for parallel workstreams via RPC.
Full web & browser
Search, extract, automate the browser, plus vision and image gen.
Plans that grow with Hermes
Every plan is a dedicated box with root, static IPv4 + IPv6, a subdomain, and all chat gateways. Monthly and annual plans — cancel any time.
Kick the tyres. A personal Hermes that's always on, for light daily use.
billed $384/yr
- 2 vCPUx86
- 4 GB RAM
- 40 GB SSD
The daily driver. Memory, skills, sub-agents and every chat app, comfortably.
billed $564/yr
- 4 vCPUx86
- 8 GB RAM
- 80 GB SSD
Heavy lifting — parallel sub-agents, batch trajectories and MLOps workloads.
billed $1236/yr
- 8 vCPUx86
- 16 GB RAM
- 160 GB SSD
Every Hermes deploy includes
- Full root SSH access
- Dedicated IPv4 + IPv6
- Subdomain + auto TLS
- All 5 chat gateways
- Persistent memory volume
- Voice-memo transcription
- Auto-updates & daily backups
- Bring your own model key
Pantheon
Fleets of Hermes agents for your whole team — bigger boxes (or GPUs), SSO, private networking, audit logs, an SLA, and a human on call.
Reply within one business day.
From plan to live agent in about a minute
No terminal required. Para carries Hermes to your new box and reports back when it lands.
Pick & configure
Choose a plan, toggle your chat apps, and paste a model key.
Hermes takes flight
Your dedicated box goes live with Hermes fully configured, all gateways wired, ready to receive messages.
It lands, ready
You get root, your IPs, and hermes.you.agentdeploy.cloud — already chatting.
Questions, answered
Do I need to set up or configure anything?
Nothing. Pick a plan, choose your chat apps, paste your model key — Hermes arrives on your dedicated box fully configured and live. No terminal, no setup wizard, no waiting.
Is it really my server, with root?
Yes. You get full root SSH on a dedicated box. SSH in, install anything, edit configs, even tweak the kernel. Hermes runs there with you, not in a locked sandbox.
Which model does Hermes use?
Your choice. Bring an OpenRouter key, any OpenAI-compatible key, or point it at a local endpoint. Olympian has the headroom to run smaller models on-box.
What happens to its memory when it updates?
Nothing — that's the point. ~/.hermes/ lives on a durable volume that survives restarts, updates, and plan upgrades, so it keeps everything it has learned.
Can I move up or down a plan later?
Anytime. Your memory volume and subdomain come with you. Plans are monthly or annual — cancel whenever you like.
Is my data private?
Hermes ships zero telemetry and all data stays on your box. We keep your box running and up to date — we never read your conversations.
Your Hermes is one click from landing.
Free trial credit to start. No card to try it. Deploy now, scale when it matters.