Billing and usage

Vesper bills what you actually use: the real model-token cost of your questions and agent runs, at a fixed margin. There are no seats and no subscription tiers.

Beta pricing

Vesper is in closed beta, and every figure on this page is a beta figure. These prices apply to the organizations testing the service today, and they will change. The team has yet to set the final pricing, so the numbers below are working figures under review rather than a published price list.

The beta bar in the console says the same thing: the pricing is work in progress and will change.

The service credit

Every organization starts with a lifetime service credit, currently $40, and no card is required. Questions and agent runs draw it down. The Billing page shows how much of it is left, on its own readout, because the credit is a lifetime figure and not a monthly one.

When the credit is exhausted and no card is on file, asking returns a clear "credit used up" state that links to the Wazuh Hub, and nothing else happens: no surprise charges, no lockout of your existing data.

What is metered

Every model call is recorded against your tenant:

• Questions: the retrieval and generation calls behind each answer.
• Corpus searches on the Answers page: the embedding call that ranks the corpus against your term. No answer is generated, so this is a small fraction of what a question costs, but it is a real model call and is metered like one. The page prints the measured figure under each result set.
• Agent runs: the tool-use loop's accumulated token usage, recorded when the run finishes or suspends for approval.

Longer conversations cost more, because recent turns are replayed into follow-up runs so the agent keeps context. That is also why the replay window is bounded.

Turning on Pay as you go

Nothing is ever charged until you switch it on. Your organization's card lives in the Wazuh Hub and is shared with every Wazuh Labs service, so a card added to keep another service running does not start billing here.

On the Billing page, a billing admin turns on Pay as you go. Until then the workspace runs on its free credit, and when that credit is gone new questions and agent runs pause rather than being charged.

You can turn it off again at any time. Billing stops at the end of the current period, and usage already recorded is still invoiced. Nothing is deleted.

If your organization has an arrangement with us that means it is not billed, there is nothing to switch on and the page says so.

Paying past the credit

Wazuh Labs services share one billing relationship per organization, managed in the Wazuh Hub:

1. An admin adds the organization card once, in the Hub.
2. A billing admin turns Pay as you go on in Vesper. The card is what makes that possible, and switching it on is what starts the billing: one added for another Wazuh Labs service never starts charging here on its own.
3. Usage is reported daily to the shared organization account, denominated in actual usage, and itemized per service, so Vesper's line is visible next to the other Labs services you use.

Your monthly budget

Metered billing does not mean unlimited billing. Every workspace has a monthly budget: a ceiling on what it can spend in a calendar month. New workspaces start at $25 per month.

• The Billing page shows this month's spend against the limit on its first readout, and warns you at 85% and 90%.
• At 100% new questions and agent runs are paused until the next billing month. Nothing is deleted and reading stays open: your conversations, history and settings are all still there.
• A billing admin can change the limit at any time, anywhere between $1 and $100 per month. Raising it takes effect on the very next question, with nothing to wait for.
• Every member of the workspace can read the limit, so somebody whose question was paused can see why.
• Need more than $100 per month? Contact us and we will raise it for you.

The budget and the service credit are separate meters. While you are inside the free credit you are not being billed at all, so in practice the budget only starts to matter once your organization has a card on file.

Keeping an eye on spend

• The Usage page is the detailed view, over 24 hours, 7 days, 30 days or all time. It counts what you did (questions asked, agent runs and how many are waiting on an approval, how many were look only, commands proposed, approved, denied and run without asking) and breaks the tokens down by feature: questions, agent runs and corpus searches, each with its own share of the spend. A billing admin also sees the money column; every member sees the tokens.
• On that table, model calls counts calls to the model, not questions or runs. One question makes up to two, one for the answer and one for the retrieval. Cached tokens are a part of input, so the two columns are never added together. The token total on the Billing page is input plus output.
• The Billing page is the summary. Three readouts across the top say what has been spent this month against the budget, how much free credit is left, and what is heading for the next invoice. Below them a spend chart shows the same by-feature split as money, as tokens, or both at once, over 24 hours, 7 days, 30 days or all time, beside the monthly budget, the Pay as you go switch, invoices and the state of the organization card.
• The chart has no per-day curve, and that is deliberate rather than missing: spend is recorded per call and reported per feature, so a line of money per day would be a monthly total divided by thirty rather than anything that happened.
• Neither page names the underlying models. The split you see is by what you used Vesper for, because which model answers a given question is ours to choose and changes as the platform improves.
• Admins can watch the Changes ledger to correlate agent activity with usage.