Your AI Agent. Always On. Everywhere You Chat.

Mercury runs your Hermes agent 24/7 on a dedicated cloud server — not your laptop. Telegram, Discord, Slack, WhatsApp. Your agent is always online, always remembering, and it never goes to sleep when you close your computer. You bring your API key. We handle everything else. From $12/mo.

See Plans & PricingSee What It Can Do
Join the first wave of early adopters running their AI agent around the clock — not just when their laptop is on.

What your agent can do

Everything your agent can do

Remembers your business

Your agent learns how you operate — your tone, your customers, your workflows — and gets smarter the longer it runs. Because it lives on a server, not your laptop, it never loses context when you shut down for the night.

Talks where you already talk

Telegram, Discord, Slack, WhatsApp — your agent lives in the apps you already use. Message it from your phone, your desktop, anywhere. Your agent is always online to receive those messages, because it's running on a server 24/7.

Works while you sleep

Scheduled reports, automatic replies, daily briefings — it runs unattended in the background, 24 hours a day. This only works because your agent is on a reliable cloud server, not a laptop that goes to sleep when you do.

Get started in
2 minutes

1

Pick your plan

Choose Starter ($12), Pro ($20), or Business ($40). Cancel anytime. No lock-in.

2

Connect your chat

We send you a bot invite for Telegram or Discord. You'll add your API key in one simple flow — no terminal, no config files. Connected in 30 seconds.

3

Start talking

Ask questions, set crons, scrape sites, track stocks. Your agent is always online, always ready. It gets smarter over time.

One agent.
Many jobs.

Web Scraping

Send any URL, get the content. Handles JavaScript sites. Runs on schedule even when you're offline.

Cron Scheduling

Set recurring tasks in natural language. 'Scrape this every morning at 8 AM.' Runs reliably because your agent doesn't need your computer to be on.

YouTube Pro

Drop a link. Get key points, a tweet, or a full writeup, delivered to your chat while you're away.

Stock Tracker

Monitor your portfolio, daily P&L, price alerts. Get morning briefings without opening an app.

Content Monitor

RSS feeds, blogs, news — summarized to your chat. New posts detected and delivered while you sleep.

Research & Web

Ask anything. Get a summary, not 10 browser tabs.

Self-hosting is free. Mercury is for people who want it to actually work.

Your laptop sleeps. Your agent dies.

A locally running agent is only available when your computer is on, awake, and connected to WiFi. Mercury runs on a dedicated cloud server — always on, always connected.

Downtime loses context.

Hermes builds a memory of you over time. If your local setup crashes, you lose that context. Mercury backs up your agent's memory and keeps it running.

Setup is not free.

Self-hosting Hermes means reading docs, managing dependencies, handling updates, dealing with errors. Mercury handles all of that. Your first message happens in 2 minutes, not 2 hours.

Your agent is a phone contact, not a local process.

You message your agent from Telegram or Discord on your phone. Your agent needs to be on a server to receive those messages when you're not at your desk.

Simple, honest pricing

You use your own API key — you control your model choice and spending. We host and maintain everything else. No markup on LLM costs, no surprise fees.

Starter

$12/mo

(~HK$93)

1 agent · Telegram + Discord · Unlimited cron · All core skills · Shared VPS container

Best for: Solo operators, freelancers

Get Started
Most Popular

Pro

$20/mo

(~HK$156)

All platforms · Unlimited cron · Web scraping · Subagent mode · All skills · 4 GB dedicated VPS

Best for: Small teams, power users

Get Started

Business

Contact Sales

Up to 5 users · Shared memory · Dedicated 8 GB VPS · SLA · Custom integrations · Priority support

Best for: Larger businesses, teams

Coming Soon

Bring your own API key (required). Cancel anytime.

Frequently Asked Questions

Why can't I just run Hermes on my own computer?
You can. But your agent will only work when your computer is on and connected. No crons while you sleep, no messages received when your laptop is closed, no memory persistence if your setup crashes. Mercury keeps your agent running 24/7 on a real server.
Do I need to know how to code?
Not at all. You describe what you want in plain English. We handle everything else.
Where does the agent actually run?
On our secure cloud VPS. You don't need to install anything or manage any infrastructure.
Which platforms are supported?
Telegram, Discord, Slack, WhatsApp. Your agent is a contact in your phone, not a website.
How do I set up my API key?
After subscribing, you provide your OpenRouter, Anthropic, or OpenAI key. We use it to power your agent. You can change it anytime from your settings. You keep full control over which model you use and what you spend.
How much will LLM API calls cost me?
That depends on your usage. A typical user spending ~30 minutes/day interacting with the agent might spend $5-15/mo on API calls through OpenRouter or a similar provider. You have full control over which model you use and how much you spend.
Is my data safe?
Yes. Your API key is encrypted and stored securely. We never share your data with third parties.
How long does setup take?
Under 2 minutes. Pick a plan, connect your chat, and start talking.
What is the difference from ChatGPT?
ChatGPT is a chatbot you open in a browser. Mercury is an always-on agent that lives in your chat apps, runs scheduled tasks, scrapes the web, and remembers you across days. ChatGPT doesn't send you morning briefings while you sleep.
What happens if I stop paying?
Your agent is paused and data retained for 30 days. Resubscribe to resume where you left off.
Can I cancel anytime?
Yes. Monthly subscription, no lock-in, no hidden fees.

Ready to have your own 24/7 AI agent?

From $12/mo, your API key, cancel anytime. No laptop required.

Get Started

No lock-in · No hidden fees · Connected in 2 minutes · You control your API key and model · Runs on a server, not your laptop