Your own AI agent, on your own server
Hermes Agent pre-installed on VPS — ready in 10 minutes. Persistent memory, 40+ tools, Telegram/WhatsApp/Discord. Your data stays with you.
Connect from anywhere
Why do you need a self-hosted AI agent?
Your data, your control
ChatGPT and Claude store conversations on their servers. With Hermes on your VPS, nothing leaves your machine. GDPR-compliant by design.
Learns and adapts
Unlike regular chatbots, Hermes remembers past conversations, learns from mistakes, and becomes more useful every day.
Predictable cost
No billing surprises. You only pay for the VPS (from €9.99/mo) — no tokens, no message limits, no OpenAI subscription.
Why self-hosted beats the cloud
ChatGPT / Claude (cloud)
- Data on someone else's servers
- Monthly subscription $20-200+
- Zero memory between sessions
- No customization
- Message / token limits
- History periodically deleted
Hermes Agent (your server)
- Data 100% on your machine
- Fixed VPS from €9.99/mo
- Persistent memory that grows
- Fully customizable (tools, personality)
- Unlimited — it's your server
- All history yours, permanently
What can Hermes Agent do?
Multi-platform
Telegram, Discord, Slack, WhatsApp, Signal, Email — all from a single gateway.
Persistent memory
Remembers who you are, what you discussed, your preferences. Gets better the more you use it.
40+ built-in tools
Execute code, search the web, manage files, send emails, schedule tasks — no extra plugins needed.
20+ LLM models
OpenAI, Anthropic, DeepSeek, Ollama (local), OpenRouter — switch models with a single command.
Enterprise security
Non-root user, UFW firewall, fail2ban, systemd hardening — same standards we apply to our own servers.
Always-on (daemon)
Not a session chatbot. Hermes runs 24/7, executes scheduled tasks, sends proactive notifications.
Plans & Pricing
EUR • excl. VAT
Agent Starter
Cloud API — ideal to start
- 2 vCPU, 4 GB RAM, 40 GB NVMe
- Hermes Agent pre-installed
- DeepSeek / OpenRouter (cloud)
- Telegram + Discord gateway
- Persistent memory
- Email support
Agent Pro
Local LLM — zero API cost
- 4 vCPU, 8 GB RAM, 200 GB SSD
- Hermes + Ollama pre-installed
- Llama 3.1 8B local (no API)
- All gateways (6 platforms)
- Persistent memory + search
- Priority support
Agent Business
Maximum power, large local model
- 8 vCPU, 16 GB RAM, 400 GB SSD
- Hermes + Ollama + 70B model
- Fully local inference
- Parallel subagents
- Training data generation
- Dedicated support + setup call
Frequently Asked Questions
Do I need to be a developer to use Hermes?
No. You interact with the agent in plain English via Telegram, WhatsApp, or terminal. We deliver the server ready to go — you just talk to it.
Which AI model does it use?
Your choice: DeepSeek (cheapest, included in Starter), Llama 3.1 local (included in Pro, zero API cost), or OpenAI/Anthropic/any model via OpenRouter.
Can I connect my business WhatsApp?
Yes. Hermes supports WhatsApp, Telegram, Discord, Slack, Signal, and email. Setup takes 5 minutes.
Is my data safe?
Everything stays on your server. Zero telemetry, zero tracking, zero cloud lock-in. Servers in the EU, GDPR-compliant.
Can I cancel anytime?
Yes, no penalty. Data on your server remains yours — download or migrate it anytime.
Can I run multiple agents on the same server?
Yes, on Pro and Business plans you can run multiple agents with different personalities and purposes.
Ready for your own AI agent?
Order now, your server is ready in 10 minutes. No contracts, no surprises.
Choose your planHermes Agent is developed by Nous Research (MIT License). EastWeb provides hosting, installation, and technical support.