Plan comparison
Four tiers. Each one buys strictly more of every dimension than the one below it, and every figure here is enforced by the engine rather than merely advertised.
| Free | Build | Scale | Archive | |
|---|---|---|---|---|
| Price | $0/mo | $29/mo | $199/mo | $799/mo |
| Credits / month | 1,000 | 50,000 | 500,000 | 2,000,000 |
| Credits / minute | 120 | 600 | 3,000 | 12,000 |
| Archive depth | 30 days | 1 years | 5 years | Full archive |
| Watches | 2 | 25 | 200 | 1,000 |
| Shortest watch interval | 60 minutes | 15 minutes | 5 minutes | 1 minute |
| At the allocation | Stops | Bills overage | Bills overage | Bills overage |
| Overage / 1,000 credits | — | $1.00 | $0.60 | $0.40 |
Prices and packaging live on the pricing page; plans are changed in the console. This page exists so the numbers your code will meet are documented next to the behaviour they cause.
Which dimension binds first
- A product with steady traffic is usually bound by the
monthly credits. Ask for bigger pages, prefer
/stories, and use keyword ranking where it answers; each cuts credits without cutting results. - A batch job or a backfill is bound by the credits a minute. Serialise with a small concurrency limit rather than bursting and retrying.
- A monitor is bound by the watch cap and the shortest watch interval. Each run of a watch costs its query's credits, and every webhook delivery attempt 1 more, so the interval floor also bounds what a watch can spend.
- Research and analysis is bound by archive depth, and that is the one no amount of client-side cleverness works around.
The free tier
A real allocation, hard-capped, with no card. It stops at its credits rather than billing you, and it reaches back 30 days — enough to build and test an integration properly, and not enough to run an archive study on.