Pay for the server, not the software.

Yes, including Free. There is no seat quota in the plan catalogue and no counter behind one, so there is nothing to hit. We meter what your workflows produce — executions, log lines, events — and an extra teammate reading the same data costs us approximately nothing.

No. There is no clock, no card, and no feature switched off to make you upgrade. All ten detectors run on it. What is smaller is the scale: one instance, a shorter history, and a monthly volume cap.

One connected n8n. Staging and production are two. A workflow count is never part of the price.

Not in agent mode — reduction happens on your host and only metadata is sent. In API-key mode, execution data transits our infrastructure before being reduced, though it is never stored. This is covered properly on the security page.

One finished n8n run, however many nodes it has. Heartbeats and the workflow snapshots the agent sends to detect missed runs are not counted — they are how the monitoring works, not something to bill for.

For a countable limit — instances, alert channels, API keys — the action that would exceed it is refused, with a message naming the limit and the plan that lifts it. Nothing already running is cut off and no data is deleted.

For a monthly volume limit, new log lines and business events are refused until the meter resets. Run monitoring never stops. Executions, detection and alerts keep working over the limit, because an observability tool that goes quiet when you exceed a quota is indistinguishable from the outage it is supposed to be telling you about.

Cancel any time and you keep access until the period you have paid for ends. If Obsevo did not do what this site says it does, email us and we will refund you. Full terms are on the refunds page.