Skip to main content
It’s the middle of the month and your credit balance is lower than you expected. Which agent is responsible — the busy support bot, or the new one you embedded last week? The Usage dashboard answers exactly that: it shows credit spend across your whole organization, broken down by agent and over time, so you can find the cause instead of guessing. Open it from Usage in the dashboard sidebar (alongside Settings). Its subtitle says it plainly: “Monitor credit usage across your agents and team.”
The Usage dashboard with the All Agents filter, date range picker, the Credits used, Chatbots used, and Team members used cards, and the Credit usage chart.

​
What’s on the dashboard

Usage is read-only — it reports what’s already happened. You’ll see a set of summary cards and a trend chart. Only the chart (and the total shown above it) respond to the agent and date filters at the top — the summary cards always show your current plan-period figures, no matter which agent or date range you pick.

​
Summary cards

These three cards reflect your current plan period and are not affected by the date range or agent filter — only the chart below reacts to those. Credits used here is spend — how many credits have been consumed, not your remaining balance. To see what’s left and when it refills, open Billing (see Credits explained).
Usage breaks credit spend down by agent and over time — there’s no per-team-member spend breakdown. Team members used is a seat count (how many of your seats are in use), not a measure of who spent what.

​
The Credit usage chart

The Credit usage chart plots credit consumption over the selected date range, so you can see spikes (a launch, a viral day) against quiet stretches. It’s a single credit-spend trend line — there’s no chart toggle. The total above the chart shows both numbers for the selected range as a headline figure — for example, “X credits · Y messages”. Reading them together tells you whether a spend spike came from more conversations or from a pricier model answering them.

​
Filter to find the cause

1

Narrow to one agent

Use the agent filter at the top — All Agents by default — and choose Select Agent to isolate a single agent’s credit usage. This is how you confirm which agent is driving spend.
2

Pick a date range

Use the date range picker (Start date – End date) to focus on a billing period, a launch week, or a custom window. The chart and its headline total update to match — the summary cards above stay on your current plan period.
Usage dashboard with an agent filter and date range, summary rings, and a populated Credit usage chart
3

Read the trend

Follow the credit-spend line for spikes, and check the “X credits · Y messages” total above it to see whether a spike is volume-driven or model-driven.

​
Options reference

​
Usage vs. Billing — what’s different

The Usage dashboard and the Billing page show related but distinct things. Don’t redraft your balance math from Usage — read it on Billing:
If one agent is consistently the biggest spender, check the model it uses on Model & settings. A lighter model can cut per-message cost substantially while still answering well from a strong knowledge base.

​
Troubleshooting

Usage only appears once your live, embedded agents have answered messages in the selected range. Testing in the Playground doesn’t draw credits, so it won’t show here. Widen the date range or confirm the agent is public and embedded.
A temporary hiccup can stop a range from loading. Re-apply the date range or reload the page to retry. If it persists, contact support@a2v2.ai.
They measure different things. Usage shows credits spent; Billing shows credits remaining. Your balance also depends on plan refills and any top-ups — see Credits explained.

Agent Analytics (Beta)

Go beyond cost — see how a single agent is performing.

Credits explained

The two credit pools, what spends them, and what happens at zero.

Auto-recharge

Top up automatically so a busy spell can’t take agents offline.

Plans

What each plan includes and how many credits it gives you.