How Qoren works

Qoren has a handful of building blocks. Once you know what each one is for, every screen in the console makes sense. This page walks through them in the order you meet them.

Your account#

When you create an account, Qoren sets up a private space for it, with you as its owner (the console calls this step "Setting up your workspace"). Everything else belongs to your account: your plan, your credit balance, your environments and agents, and your business details.

The business details matter more than they look. The name of your business, what it does and how it makes money are handed to every agent you hire, so no agent has to ask you again. You fill them in once, in the welcome dialog or on Company details.

Environments#

An environment is a private computer in the cloud that your agents live and work on. Qoren creates it, keeps it online and installs what your agents need. You never have to look after a server.

Environment sizes are named by how many agents they hold, from 1 agent up to 10. Several agents can share one environment, and each runs under its own separate account on it. An environment draws credits by the hour while it runs, whether or not its agents are busy. Read more in What an environment is and Environment sizes.

Agents#

An agent is one AI worker with one job, such as writing a morning brief, answering enquiries or chasing invoices. It lives on one of your environments and has everything it needs to do that job:

• a runtime, the agent software it runs on
• a model, the AI that does its thinking, which you can change later
• its instructions: who it is and how it works, in the Brain section of its settings (see Identity and instructions)
• the keys its tools need, such as an API key for your CRM
• connections to chat apps, other tools and, if you like, its own mailbox
• a workspace: the files it works with

Each agent has its own page in the console, with tabs for chatting, its activity, its files and its settings. To make your first one, follow Deploy your first agent.

Runtimes#

The runtime is the agent software underneath. Qoren runs three: Hermes, OpenClaw and Codex. A template says which runtimes it is built for and picks one for you, so most people never have to think about it. See Runtimes for how they differ.

Templates and skills#

A template is a ready-made agent: its instructions, the tasks it runs on a schedule, the files it starts with, and the runtimes it supports. Pick one when you create an agent, or save your own. See Templates.

A skill is a reusable set of instructions you can add to any agent. Write it once, then add it to every agent that needs it. See Skills.

The Vault#

The Vault holds the API keys and passwords your agents use, encrypted. When a template needs a key, you can pick it from the Vault instead of typing it again, and you choose which agents can use which key. See Keys and passwords.

Connections#

An agent does not have to wait for you to open the console. You can:

• reach it on Telegram, Slack or Discord
• give it scheduled tasks that run on their own
• let other tools wake it with an event, through a webhook
• give it an email address of its own

Credits#

Your account has one credit balance, and your plan adds to it every month. Everything that costs money draws from it: environments by the hour, the AI model's work when it runs on Qoren's model key, and web searches. If you bring your own model key, you pay the model provider directly for that part.

You can top up at any time and set a monthly budget. If the balance runs out, your agents pause until you add credits, and your environments stay in place. See How credits work.

Where you control it all#

Most people use the console at qoren.sh. Everything you do there goes through the same Qoren API that the command line tool and the SDK use, so you can script the same work later.