Choose the AI model an agent uses

The model is the AI that does the thinking for an agent: it reads each message, decides what to do and writes the reply. You pick one when you create the agent, and you can change it at any time from the agent's settings, along with its reasoning effort: how hard the model thinks before it acts. A stronger model or a higher effort handles harder work; a lighter one keeps usage down.

Change the model or the effort#

1. In the sidebar, click Agents and open the agent.
2. Click the Settings tab. It opens on General.
3. The AI model row shows the model the agent runs on and its provider.

12

• To change the effort, click Low, Medium or High on the effort switch (1). It saves straight away.
• To change the model, click Change (2). The Choose a model dialog opens.

In the dialog:

1. Narrow the list with the filters on the left (1), search by name, ID or description (2), or reorder it with the sort menu (3).
2. Click a model to select it. If you want a different effort as well, pick it on the switch at the bottom (4).
3. Click Save (5). Double-clicking a model saves it with the effort shown.

12345

While a change is applied the row says "Saving". When it finishes it says Saved, and the agent uses the new model and effort from its next run.

Which effort to pick#

• Low answers fastest and uses the fewest tokens. Good for routine replies, sorting and short lookups.
• Medium suits most agents: it thinks through multi-step work without spending much more.
• High takes longer and uses more tokens on each turn. Use it for hard problems such as research, planning or code.

Models that do not reason ignore the effort, so the setting changes nothing for them. An agent made from a template starts on the effort the template chose.

What the model list shows#

Each model shows its name, its ID in the form provider/model (such as anthropic/claude-sonnet-4-6, where the part before the slash is the company that makes the model), a short description, how much text it can read at once (its context window), and when it was added to the catalog. On the right are its input and output prices.

• Prices are the provider's list prices from OpenRouter, in US dollars per million tokens. Input is what the agent sends the model; output is what the model writes back. On the managed key, calls are charged in credits based on these prices; on your own key, your provider bills you directly.
• Recommended: models Qoren has marked as a good fit carry a Recommended tag and sort to the top by default. Turn on Recommended only to see just those. They are a suggestion, not a limit: every model in the list is available on every plan.
• Sorting: by name, price (cheapest or priciest first, by input plus output), largest context window or newest.

Filter the list#

The sidebar narrows the list as you change it. On a phone, open it with Filters; the number on the button counts the filters in use.

• Input price and Output price: drag the two handles of each slider to set the lowest and highest price you want, or type exact dollar figures into the boxes under it. The slider spends most of its length on the cheaper end, where most models are. A handle at the far right means no upper limit.
• Released: drag the handles to a range of months, or pick Past 3 months, Past 6 months or Past year.
• Context window: show only models that can read at least 32K, 128K, 200K, 400K or 1M tokens at once.
• Provider: tick one or more companies. The number beside each is how many of its models are listed.

Reset at the top of the sidebar clears the filters and keeps your search. Clear search and filters above the list clears both. A model with no listed price or release date, such as an ID your team saved itself, is hidden while a price or date range is set.

• The current model is always listed and tagged Current, even if it has since left the catalog, so you never lose it by opening the picker.
• The list is the same for every runtime. Hermes, OpenClaw, Codex and OpenCode agents choose from the same catalog.

There is no box for typing a model ID by hand here: you pick from the list.

How to choose#

• Start with a recommended model and change only if the agent struggles or costs more than you want.
• Pick a stronger model for agents that plan several steps, write long documents or work with code.
• Pick a lighter model for agents that sort, summarize or reply to routine messages all day.
• After a change, watch the agent's Usage tab for a few days to see what the switch did to its spend.

Managed key or your own key#

Every model call is paid for in one of two ways, set per agent:

• Managed key (the default). Qoren provides the model key and charges the agent's model calls in credits from your balance. See how credits work for what a call costs.
• Your own key. If you add your own provider key (OpenRouter, Anthropic, OpenAI and others) under Settings, BYOK, you can put one agent on it from its Model key setting, or all of them by turning BYOK on. Model calls on your key go to your provider, which bills you directly. Qoren does not charge credits for those calls. Web search and mailboxes are still charged in credits either way. Setup is in bring your own key.

Switching between the two does not change which model an agent uses. It only changes who pays for the calls.

The Runtime row#

Runtime shows the agent software the agent runs on: Hermes, OpenClaw, Codex or OpenCode. It is chosen when the agent is created and cannot be changed from this page. The runtime decides some features, such as whether the agent can import skills. Runtimes compares them.

The Health row#

Health at the top of General is a one-line answer to "is this agent all right?". Checks run on their own every few minutes, and the row says when the last one ran. If the agent is running, you can click Check now to run one straight away.

When the status is red, repair the agent. When it says Paused, see spend controls.