Why Self-Host Hermes?

An agent that lives on your server, remembers everything, and gets smarter every day. 184k stars on GitHub. MIT license.

Persistent Cross-Session Memory

Hermes remembers your preferences, projects, and environment across every session. The longer it runs, the better it knows you — no re-explaining context.

Automated Skill Creation

When Hermes solves a hard problem, it writes a reusable skill document so it never forgets how. Skills are searchable, shareable, and follow the agentskills.io standard.

Multi-Platform Gateway

Connect Telegram, Discord, Slack, WhatsApp, and Signal — all from a single gateway process. Start a conversation on Telegram, pick it up in your terminal.

Scheduled Automations (Cron)

Built-in cron scheduler with delivery to any platform. Set up daily reports, nightly backups, weekly audits — all running unattended on your VPS.

Parallel Sub-Agents

Spawn isolated sub-agents for parallel workstreams. Each gets its own conversation and terminal. Collapse multi-step pipelines into zero-context-cost turns.

100% Self-Hosted & Private

All data stays on your VPS in ~/.hermes/. Zero telemetry, zero tracking, MIT license. You own your agent, your data, and your workflows.

Install in One Command

SSH into your VPS and run the official installer. No prerequisites needed.

# Install Hermes Agent (handles everything automatically):

curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash

# Configure your LLM provider:

hermes setup

# Start the messaging gateway (Telegram, Discord, etc):

hermes gateway

Choose Your VPS Plan

Recommended configurations for running Hermes Agent 24/7

All Plans Include: Full Root Access Dedicated IPv4 NVMe SSD 24/7 Support
Starter

VPS 1

For light agent tasks

55% OFF
रु899/mo
with 12-month term
  • 4 vCPU Cores
  • 6 GB RAM
  • 100 GB NVMe SSD
  • 30 TB Traffic
Get Started
Best Value
Recommended

VPS 2

Ideal for Hermes + messaging gateway

36% OFF
रु1,900/mo
with 12-month term
  • 6 vCPU Cores
  • 16 GB RAM
  • 200 GB NVMe SSD
  • 30 TB Traffic
Get Started
Business

VPS 3

Multi-platform + sub-agents

रु4,167/mo
with yearly term
  • 8 vCPU Cores
  • 24 GB RAM
  • 300 GB NVMe SSD
  • 30 TB Traffic
Get Started
Most Power
Enterprise

Cloud VPS 4

Heavy workloads + local LLMs

रु5,000/mo
with yearly term
  • 12 vCPU Cores
  • 48 GB RAM
  • 400 GB NVMe SSD
  • 30 TB Traffic
Get Started

Frequently Asked Questions

Hermes Agent is an open-source, self-improving AI agent built by Nous Research (184k GitHub stars). It runs on your own server with persistent memory, automated skill creation, multi-platform messaging (Telegram, Discord, Slack, WhatsApp, Signal), and scheduled automations — all under an MIT license with zero tracking.
Hermes needs Python 3.11+, Node.js, and at least 4 GB RAM for basic use. For production with messaging gateway + sub-agents + memory, we recommend VPS 2 (16 GB RAM, 6 vCPUs). The installer handles all dependencies automatically.
Hermes works with Nous Portal (native OAuth), OpenRouter (200+ models), OpenAI, Anthropic Claude, NVIDIA NIM, Hugging Face, local vLLM, or any OpenAI-compatible endpoint. Switch providers with `hermes model` — no code changes.
Hermes is a persistent autonomous agent that lives on your server 24/7. It remembers across sessions, creates reusable skills from experience, runs scheduled tasks, spawns sub-agents for parallel work, and reaches you on Telegram/Discord/Slack. It grows smarter the longer it runs.
One command: `curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash` — this installs uv, Python 3.11, clones the repo, and sets up everything. Then run `hermes setup` to configure your LLM provider and `hermes gateway` for messaging.