Every endpoint in the catalog has a real price, not a credit abstraction. The price is worked
out before the call runs, from the tool’s price shape and the input you gave it.
Discovery and inspection are free and need no account. Only running an endpoint spends.
The price shapes
| Shape | How it prices | Example |
|---|
| Per call | A fixed rate, whatever comes back | One company lookup |
| Per result | Rate × the number of rows you asked for | 50 search results |
| Per unit of time | Cannot be priced from the input — refused | Audio or video billed by duration |
| Metered | Cannot be priced from the input — refused | Anything billed by token count |
A per-result tool is priced at its worst case: the limit you asked for, assumed to be
reached. Ask for 100 rows and you are quoted 100 rows.
An estimate is a worst case, not a guess
The number quoted before a call is the most it can cost, not the most likely cost. That
is deliberate — a cap has to be checked against something that cannot be exceeded.
So the two numbers you will see on a call rarely match:
- Quoted — the worst case, computed before anything was sent.
- Settled — what it actually cost, from what actually came back.
A per-result tool that returns 12 rows when you asked for 50 settles on 12. A per-call
tool settles at its rate. A call the vendor failed settles at zero — a vendor’s bad
day never lands on your bill.
Dry runs
A dry run prices a call and shows you exactly what would be sent, without sending it.
It costs nothing. Use it when:
- The tool is new to you and you want to see the request before it goes.
- You are about to run something across a list and want the per-item price first.
- A call was refused and you want to see which number tripped the cap.
Ask for it in plain language in chat, or set dryRun on an HTTP run.
Price this before you run it — how much for 200 results?
Where the number on your bill comes from
Every attempt is recorded: tool, vendor, status, quoted price, settled cost. That record
is what your balance is drawn against and what the daily cap is measured against. It is
the source of truth, not a summary of one.
Read it on the dashboard: one row per call, with the full run
detail behind each — quoted price, settled cost, status and which endpoint ran.
Plans and prices
Endpoint prices are per call and are the same whatever plan you are on. A plan changes
how many credits you get toward them each cycle, not what a call costs. One credit is
$0.01 of catalog price, and the ledger is finer than a credit — most endpoints cost a
fraction of one, and prices are never rounded up to a whole credit.
Plan details and prices: https://agentmore.app/pricing.
Next: Spend caps for how to bound what a call may cost, and
Plans & billing for how credits arrive. 
