ScaleMax Cloud
Launch

Hermes Agent · Agent hosting

Hermes Agent hosting, ready in one click

Host Hermes Agent, the self-improving AI agent by Nous Research, on a KVM server that stays on. It keeps its memory between sessions, runs with the model you choose and answers you on Telegram or WhatsApp.

Launch for ₹599/moSee all plans

What you get

An agent that remembers

  • Installs by itself

    Pick it at checkout and it installs as soon as your server is ready, in a few minutes, then starts by itself.

  • Memory that persists

    Memory, skills and settings live on your server’s disk, and scheduled backups copy them offsite to Finland (EU).

  • No extra password

    Hermes Agent’s dashboard opens on your server page with its login handled for you.

  • Any AI model you choose

    Paste a ScaleMax key (many models on one balance, from the same team; it is checked and a model picked for you), or use your own OpenAI, Anthropic or OpenRouter key, or any public https endpoint that speaks the OpenAI or Anthropic API.

  • Telegram and WhatsApp

    Connect a Telegram bot or link WhatsApp from the dashboard. Only you can chat with it; group chats are off.

  • Light on memory

    Hermes Agent idles around 239 MB of RAM, so the smallest plan is enough for API-only use.

How it works

Three steps to your first chat

  1. 01

    Pick your app

    Hermes Agent is already picked when you launch from this page. At checkout it sits under “What do you want to run?”, at the same price as a plain server.

  2. 02

    Pay

    In rupees, by UPI, card, net banking or wallet, from ₹599/mo. Your server is live in about three minutes, and Hermes Agent installs by itself right after.

  3. 03

    Chat with it on WhatsApp

    Your server page shows a short Get started list: paste a ScaleMax key or use OpenAI, Anthropic or OpenRouter, scan a QR code to link WhatsApp or add a Telegram bot, then say hi. You can also chat with it right on your server page.

Flight plans

Pick a size. Pay monthly, in rupees.

The price you see is the price you pay. No setup fee, no bandwidth bill after the fact, and the price doesn't jump when you renew. Pay for a year and save 17%.

  • KVM 1

    ₹599/mo

    or ₹5,966 a year, saving ₹1,222

    vCPU
    1
    RAM · DDR4
    4 GB
    NVMe
    50 GB
    Transfer
    4 TB/mo

    Shared IPv4 + 20 ports and your own subdomain. Own IPv4 +₹249/mo.

    Launch KVM 1

  • KVM 2

    Popular

    ₹849/mo

    or ₹8,456 a year, saving ₹1,732

    vCPU
    2
    RAM · DDR4
    8 GB
    NVMe
    100 GB
    Transfer
    8 TB/mo

    Shared IPv4 + 20 ports and your own subdomain. Own IPv4 +₹249/mo.

    Launch KVM 2

  • KVM 4

    ₹1,599/mo

    or ₹15,926 a year, saving ₹3,262

    vCPU
    4
    RAM · DDR4
    16 GB
    NVMe
    200 GB
    Transfer
    16 TB/mo

    Dedicated IPv4 included. Every port is yours.

    Launch KVM 4

  • KVM 8

    ₹2,999/mo

    or ₹29,870 a year, saving ₹6,118

    vCPU
    8
    RAM · DDR4
    32 GB
    NVMe
    400 GB
    Transfer
    32 TB/mo

    Dedicated IPv4 included. Every port is yours.

    Launch KVM 8

Every Launch button opens checkout with Hermes Agent picked. It installs by itself as soon as your server is ready, at no extra cost.

Need another size? Build a custom VPSEvery add-on and yearly price

Every plan runs on dedicated AMD EPYC cores with DDR4 ECC registered memory and mirrored NVMe.

Good to know

The limits, up front

  • Sizing: KVM 1 (1 vCPU, 4 GB RAM) at ₹599/mo is enough for API-only use; KVM 2 (2 vCPU, 8 GB RAM) at ₹849/mo leaves room for heavier tools such as a browser.
  • One app runs on a server at a time. Switching to another is free, from your server page: the new app replaces the old one and deletes its data, after you confirm. Chat apps answer only while Hermes Agent is running.
  • Model usage is billed by your model provider, or from your ScaleMax balance with a ScaleMax key.
  • The servers are in Gravelines, France (EU), and backups are copied offsite to Finland (EU). There is no India region yet.

Questions

Before you launch

Ports, backups, payments and refunds are covered on the FAQ page.

Read every answer
What is Hermes Agent?

A self-improving AI agent by Nous Research. It keeps memory between sessions and gets better at the work you give it. It runs on your server with the AI model you choose.

Do I need to know coding?

No. Pick Hermes Agent at checkout and it installs by itself. Your server page then shows each next step in plain words: connect an AI with a key, scan a QR code to link WhatsApp or add a Telegram bot, and say hi. You never need the terminal, though full root is there if you want it.

Is the smallest plan enough?

For API-only use, yes: Hermes Agent idles around 239 MB of RAM. KVM 2 (2 vCPU, 8 GB RAM) at ₹849/mo leaves room for heavier tools.

Where is the agent’s memory kept?

On your server’s disk, in /opt/hermes-data. Scheduled backups copy the disk offsite to Finland (EU). Switching to another app deletes it, so the switch asks you to confirm first.

Which AI models can Hermes Agent use?

Paste a ScaleMax key (many models on one balance, from the same team; it is checked and a model picked for you), or use your own OpenAI, Anthropic or OpenRouter key, or any public https endpoint that speaks the OpenAI or Anthropic API. The key is written straight into the app on your server; the dashboard keeps only the provider, the model and the key’s last four characters.

Can I chat with it on Telegram or WhatsApp?

Yes. Connect a Telegram bot or link WhatsApp by scanning a QR code, from your server page. Only you can chat with it, and group chats are off.

Can I switch apps later?

Yes, free, from your server page. One app runs on a server at a time, so the new app replaces Hermes Agent, deleting it and all its data, and you confirm before anything goes. To keep both, use two servers.

Is AI model usage included in the plan?

No. The plan pays for the server. Model usage is billed by your model provider, or from your ScaleMax balance when you use a ScaleMax key.

Where are the servers?

The servers are in Gravelines, France (EU), and backups are copied offsite to Finland (EU). There is no India region yet. Agents and automations mostly talk to AI model APIs and websites in Europe and the US, which are close to France; typing in a terminal from India has a little more lag than a server in Mumbai would.

Keep reading

Related pages

Go for launch

Put Hermes Agent in orbit today.

Live in about three minutes. Pay monthly in rupees, cancel whenever you like, and your first order has a 7-day money-back guarantee.

Launch for ₹599/moAsk a question on WhatsApp