- Per-row price. Each API has a price per row (per profile, per post, per product), shown on its page and in the catalog. Cost is roughly
rows × price per row. - Search fees. A few search APIs add a small search fee, either once per run or per search page (LinkedIn people search: $0.05 per page of 25). It’s stated on the API’s page, and it applies even when the search finds nothing.
- Add-ons. Some APIs charge extra for optional data, such as photos on Google Maps places, only when you turn it on.
- Metered AI APIs. The
ai/*APIs bill exactly each call’s cost at the chosen model’s own rate.
What you’re not charged for
- A run that fails.
- A run that finds nothing, apart from a flat search fee where one applies.
- Rows a run didn’t deliver. A run that stops partway is charged only for what it returned.
Controlling cost
Set the API’s row-limit input (maxResults, maxItems, maxReviews and so on) to cap a run. Check your balance with GET /v1/balance before large runs, and see spend per day and per API with GET /v1/usage/daily.
Every finished job reports charged, usd_charged and billing_status, so you always know what a run cost. See Jobs and results.
Topping up
Add balance in the console. If your wallet runs out mid-run, you keep the rows collected so far and the job stops withstop_reason: "balance_exhausted".
Running at volume? Book a call for enterprise pricing.