qoren.sh

Our biggest release yet. The headline is the Office, a living 3D view of your whole fleet. Alongside it: a much larger template library, chat that lives right inside the office, richer controls for every agent, new plans, and a lot of polish throughout.

The Office: see your agents at work

Your environments now come to life as a 3D office. Each environment is its own room, sized to your plan, with every agent shown as a robot at its desk doing what it is really doing right now.

• One room per environment, with a day and night toggle and per-room decor

• Live CPU, memory, and storage readouts for the room you are focused on

• When agents in an environment collaborate, you can watch their messages fly between desks

• Slack and Telegram activity shows up in the room, and a broken connection is flagged clearly so you notice it fast

• Custom banner text for each room

• Works even on devices without 3D acceleration, thanks to an automatic fallback view

Progression and rewards

Your office and every agent now grow over time, based on the real work getting done.

• Agents earn XP and levels from completed jobs and time in service

• Classes emerge from how each agent actually works: Researcher, Operator, Scribe, Communicator, Analyst

• Commendations for milestones like a first job, a live channel, and a full room

• Your whole office earns a standing, from Startup up to Powerhouse, with a meter to the next tier

• A first-week quest board turns getting started into six clear goals

• Level-ups, rank pips, and a trophy shelf all show up live in the 3D scene

Everything is calculated honestly from real activity, so your progress always reflects what your fleet has genuinely done.

Cosmetics you earn

Avatars, rooms, and your whole office unlock new looks purely from progress, never from a purchase.

• Crests, auras, finishes, decor packs, banners, and skyline, mood, and floor themes

• A decorate mode with live preview in the 3D scene and autosave

• New accent colors and new shapes to choose from

• Locked items preview dimmed, so you can see what is coming next

Chat, right inside the office

The agent chat is now a real, movable window on the office stage.

• Drag, resize, and maximize it, and keep several agent chats open at once

• A Stop button ends the current turn instantly and keeps any partial reply

• Full formatting in replies, with one-click copy

• Long conversations stay healthy automatically, with a clear in-thread marker and a context meter

• A turn in progress keeps running even if you navigate away or reload the page

A much bigger template library

A major expansion of ready-to-deploy agent templates across engineering, business, security, and operations, including:

• Engineering: PR reviewer, ticket to PR, engineering pulse, dependency upgrader, application security reviewer, incident triager, infrastructure guardian

• Business: founder OS, investor update writer, churn autopsy, client onboarder, subscription auditor

• Operations: stock watchdog, doc keeper, applicant screener, review requester, backup verifier, meeting prepper, tax pack preparer, demand scout, bid catcher, contract redliner, ad-spend watchdog, payment rescuer

• Security: security hygiene auditor

The template designer is better too. It now runs in the background (no more two-minute limit), shows a live status timeline, offers multiple-choice answers with recommended picks, can defer questions to onboarding, and gives you a one-click fix for safety findings.

Reusable skills

• Write reusable skill docs and import them into any agent

• 20 curated starter skills with categories and previews

• Each skill shows at a glance whether it can be added, is already added, or has an update available

More you can do with each agent

• Memory: edit your agent's identity and operating manual as plain text, and manage its files

• Interactive shell: a safe command line for your agent, with sensible timeouts

• Activity view: what the agent is doing now, what is scheduled next in plain English, and a recent log

• Configuration: keys, messaging channels, and team mailbox setup

• Model picker: a curated catalog, pinned to the agent's current model

• Repair: fix a stuck agent inline, with live progress

• Rename agents from the simple interface

• Move an agent from one environment to another, safely

Environments made simple

• A simple Environments page to create, resize, and destroy environments inline

• Your agents grouped by environment, one row each

• Templates can recommend or require a minimum environment size

• Set a default region for new environments

• Each plan now includes environment capacity, with larger setups available as add-ons at checkout, and you are only charged for what goes over your included capacity

Plans and billing

• New pricing: Starter $19, Pro $49, Ultimate $129, Business $299

• Bring your own model key as a first-class option, with automatic validation and a clean switch-over

• Credit top-ups

• Explore freely: you only reach a paywall when you actually spend, not before

• A cleaner account area with tabs for Plan, Usage, Credits, and your own key

• Cancelling your account now stops billing, and reversing the cancellation restores it

Agent email inboxes (add-on)

Give an agent its own dedicated, automatically provisioned email inbox as a paid monthly add-on. Mailboxes belong to your organization, so they can be reassigned and outlive any single agent.

Faster, more reliable chat

• Agents you are actively chatting with now reply noticeably faster, with streaming responses

• Lower, more predictable latency on chat turns

• Smoother handling when the platform is briefly busy, so safe actions retry quietly instead of failing

Security and privacy hardening

We completed a broad security review and tightened access across the platform: stricter permission checks on sensitive actions, stronger isolation between organizations, email confirmation for new sign-ups with a 6-digit code, and least-privilege keys for integrations. Your data stays scoped to your organization.

Interface polish

• A calmer, refreshed look with clearer status colors, so positive states no longer show in red

• Settings reorganized into tabs: Profile, Password, Notifications, and Data and privacy

• Account menu moved to the bottom of the sidebar

• The simple interface is now mobile-friendly, with a slide-in drawer on smaller screens

• Signing in now takes you straight to your Office, with Home still one click away

• Agent actions tidied into a single menu

Learn

A free 7-lesson beginner course is now live at https://qoren.sh/learn, to help non-technical users get started in the world of AI.

Today we're introducing Qoren  v0, our first release. As the number indicates, this is a beta release.

Qoren is managed hosting for OpenClaw and Hermes agents. You deploy an agent to a dedicated cloud environment, keep it online on schedules and triggers, and watch its activity, usage, and spend from one dashboard, all without touching Docker, a VPS, runtime updates, or any of the infrastructure underneath.

This post introduces the platform and walks through what shipped in v1. If you'd rather just try it, head to qoren.sh.

---

Why we built it

OpenClaw and Hermes are amazing pieces of software. But their setup and configuration quickly becomes a devops side-quest that nobody really asked for. A server to provision, Docker to babysit, runtimes to update, secrets to store safely, model spend to watch, or a laptop that can never go offline. Most people try it, quickly get overwhelmed, and then give up.

We believe these tools can strongly benefit businesses, so long as they can actually focus on their business instead of infrastructure that should be boring and predictable.

Qoren removes that gap. You bring the agent; we run the environment. The default experience hides the infrastructure entirely, while the controls are there when you want them.

---

What shipped in v1

Deploy OpenClaw or Hermes agents

Qoren hosts two runtimes. Deploy an existing OpenClaw setup, or start a Hermes agent from a template, persona, schedule, and set of tools. If you're unsure which runtime fits, the OpenClaw vs Hermes guide walks through the decision before you commit to where you host it.

Single or multi-agent environments

Run one agent on its own, or a team of agents that work together and communicate inside the same managed environment. Higher plans allow more agents per environment, so you can grow a single assistant into a small crew that splits the work.

Fully managed environments

Every environment is a dedicated, isolated cloud box. Qoren handles provisioning, runtime setup, isolation, and per-agent resource limits so one busy agent can't starve its neighbours. You pick a size (Light, Standard, Heavy, or Max) and we take care of the rest. No region picking, no droplet wrangling, no patching. You can choose a region in specific, but you're not forced to think about it if that's irrelevant to you.

Always-on execution

Agents run on schedules and triggers without depending on your laptop, a local machine, or a self-managed VPS. Set a daily summary, a recurring check, or a trigger-driven task and let it keep working in the background.

Start fast with templates

Browse a gallery of ready-made agent templates and deploy one in a few clicks. Most agents are live within a few minutes of picking a template. You can also author your own org-scoped templates from a persona, prompt, schedule, seed files, model, and an allowlist of tools.

Your model key or ours

Use the managed model key included with your plan, or bring your own. Bring-your-own-key (BYOK) is available on Pro and above: connect your own model-provider key, pay the provider directly, and run with no managed usage cap.

Predictable usage and prepaid credits

Every plan includes a monthly LLM usage allowance. Usage draws from that allowance first, then from prepaid credits that never expire. When both run out, agents hit a hard stop, so you're never billed for usage you didn't pay for in advance. Optional auto-recharge keeps power users running without interruption.

One dashboard for everything

Activity, logs, included usage, credit balance, and spend all live in the same operational dashboard, so you always know what your agents are doing and what they're costing.

---

How it works

1. Choose the runtime. Pick OpenClaw or Hermes, from a template or your own setup.
2. Ship to a managed environment. Deploy a single agent or a collaborating team to a dedicated cloud environment. Qoren handles provisioning, isolation, runtime setup, and limits.
3. Keep it online. Run on schedules and triggers using our managed model key or your own, and monitor activity, usage, and spend from one place.

---

Plans and pricing

Every plan includes free environment capacity you can spend across one or more environments, plus a monthly LLM usage allowance shown as a multiple of the Starter allowance. Agents per environment, model access, and included usage all scale with the tier.

Included environment capacity is a value bundle, not a fixed box: Starter gets 1 Light, Pro gets 1 Standard or 2 Light, Ultimate gets 1 Heavy, 2 Standard, or 4 Light, and Business gets 1 Max, 2 Heavy, 4 Standard, or 8 Light. Extra environments beyond the included capacity are paid add-ons priced by size. Overage runs on prepaid credits that never expire.

Annual billing gives you two months free.

---

Early-bird offer

For our first subscribers, the early-bird discount takes 20% off monthly plans or 40% off annual plans, locked in for the life of your subscription. It's limited to the first 101 subscribers and only if you subscribe by July 6, 2026.

And if it's not the right fit, cancel any plan within 14 days for a no-questions-asked refund, minus any usage already spent. That applies to every plan, all the time, not just during the early-bird window.

---

Two ways to work with us

Qoren is the self-serve product: you deploy and run agents yourself. If you'd rather have it figured out and built for you, Magentic Studio also offers a done-for-you path at magentic.studio, an expert AI audit that maps where AI fits in your business, plus managed build and deployment.

Use Qoren to run agents yourself. Use the audit if you'd rather have it handled end to end.

---

Get started

Qoren is opening to new teams in batches during closed beta. Join the waitlist at qoren.sh and we'll send an invite as soon as a spot opens. Once you're in, pick a template, deploy, and your first agent is usually live within minutes.

Welcome to v1. This is the foundation, and there's a lot more coming.