One monthly budget of runs and tokens, shared across all your agents. Start free, move up when it pays for itself.
Kick the tyres and ship your first agents.
For agencies and teams running agents for clients.
For people running real automations every day.
Every plan includes the full prebuilt library, unlimited agents on paid tiers, and your data exportable at any time. No setup fee, no per-seat charge, cancel whenever.
One execution of one agent, whether it was triggered by its schedule, by you pressing run, or by an incoming webhook. A run that fails still counts, because it still did the work of trying.
Steps that use AI — reading a page and pulling structured fields out of it, scoring a CV, writing a summary. Steps that are plain code cost nothing beyond the run itself.
Some sites have no API and sit behind bot protection. Those pages are opened by a cloud browser, and each page it opens spends a credit. Agents that only call ordinary APIs never touch them.
The agent stops and tells you, rather than quietly charging you more. You can raise the plan at any point in the month and the new budget applies immediately.
No — one budget for the account, shared across every agent you run. A quiet month for one agent leaves more room for the others.
Describe the job once. It builds itself, tests itself, and gets to work.