LedgerCap

Spend-cap API

Beta

Hard caps for AI agents.
Even at 100 concurrent calls.

LedgerCap enforces a prepaid credit limit in real time using a single atomic database write — the balance never overshoots, no matter how hard agents hammer the endpoint.

Request access →Read the docs

Currently in beta — join the waitlist.

API · https://api.ledgercap.dev/api/v1

The core of it...

Charge against your account — cents or nanodollars, your call

curl -sX POST https://api.ledgercap.dev/api/v1/charge \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"amountCents":0.5,"description":"gpt-4o call"}'  # = $0.005

# {"allowed":true,"balanceNanos":4995000000,"ledgerId":"…"}
# (amountNanos:5000000 is identical — pick the unit that fits your model)

Charge a SaaS user's card — offload the whole ledger

curl -sX POST https://api.ledgercap.dev/api/v1/charge \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"amountCents":0.5,
       "externalId":"user_abc",
       "createIfMissing":true}'

# {"allowed":true,"balanceNanos":995000000}
# card auto-provisioned on first call, no pre-registration

Check the spend pool

curl -s https://api.ledgercap.dev/api/v1/balance \
  -H "Authorization: Bearer $TOKEN"

# {"availableNanos":995000000,
#  "spentTodayNanos":5000000,
#  "dailyLimitNanos":1000000000}

The second call is the offload story: same /charge endpoint, add externalId + createIfMissing: true and LedgerCap manages a separate balance, cap, and ledger for each of your end-users. No pre-provisioning step.