Loop Agent · Paid beta

Your own AI agent, running in the cloud.

Not a chat tab. A dedicated agent instance that remembers you, browses the web, runs scheduled jobs, and works while you sleep. Managed by LoopXXI.

Always on

Lives on managed cloud infrastructure, not your laptop. Works while you are away.

Real capabilities

Web browsing, file work, code execution, scheduled reports, persistent memory.

Bitcoin native

Pay over Lightning. No card on file, no subscription trap. Your instance, your term.

How it works

Live within one business day.

Each agent is a dedicated managed instance. Provisioning is confirmed by email.

1

Pick a tier and pay

Pay the Lightning invoice below, or email us for card payment.

2

Tell us where you chat

Email your order code to business@loopxxi.com with your preferred channel: Telegram, Discord, or API.

3

Meet your agent

We provision your dedicated instance and send connection details within one business day.

Pricing

Three tiers. Paid in sats.

30 days of a dedicated managed agent instance. Model usage is metered inside your tier's monthly call budget.

Starter

21,000 sats
≈ $13 / 30 days
  • Dedicated managed instance
  • 1,000 agent calls / month
  • Web browsing and memory
  • Telegram, Discord, or API access

Pro

105,000 sats
≈ $65 / 30 days
  • Everything in Starter
  • 10,000 agent calls / month
  • Stronger default models
  • Priority provisioning

Dedicated

525,000 sats
≈ $325 / 30 days
  • Unlimited calls, fair use
  • Custom tools and skills
  • Custom model routing
  • Direct support line

Pay this Lightning invoice

Your order code: —
After paying, email this code and your preferred chat channel to business@loopxxi.com.

Awaiting payment confirmation…

Prefer card payment? Email business@loopxxi.com and we will send a Stripe invoice.

What Loop Agent is, honestly

  • Each instance runs the open-source Hermes Agent runtime (MIT license, built by Nous Research), deployed and managed by Loop XXI on isolated cloud infrastructure. LoopXXI is not affiliated with Nous Research.
  • Model usage is metered through Loop XXI's gateway inside your tier budget. Conversations happen in your channel; we do not read them in normal operation, and instance memory belongs to your instance.
  • Beta terms: provisioning within one business day, 30-day term per payment, no auto-renew. If we cannot provision your instance, the payment is refunded in full.
  • No high-stakes autonomy: agents are not configured to move money, log into your accounts, or act on high-stakes matters without your explicit per-task confirmation.