Runs and allowances
Every plan is a fixed monthly price for a fixed allowance. Nothing is metered: the workspace never shows you a bill for model tokens, and a run never costs more because a site was large.
What the allowance is
Section titled “What the allowance is”| Per site, per month | |
|---|---|
| Full runs | 1 on Single site and Portfolio; 4 on Enterprise |
| Change runs | 4 on Single site and Portfolio; 16 on Enterprise |
| Pages per run | 1,000 · 2,500 · 10,000 |
Allowances are per site: on a twelve-domain portfolio, each domain has its own full and change runs. They reset on the first of each month. Usage in the workspace shows what each site has used.
Full run and change run
Section titled “Full run and change run”A full run samples up to the page budget across every template and language, fetches each page as every machine identity, builds the knowledge graph, simulates retrieval and writes the analysis. It is the monthly report.
A change run re-fetches exactly the pages the last full run sampled — the same URLs, so every difference is a change on the site and not a change of sample — plus a small slice (10% of the budget) of URLs the sitemap has listed since. It re-checks access, technical facts and performance, and reports what moved: new, resolved and regressed findings, pages whose content changed or stopped returning 200, and pages that AI crawlers can newly read or are newly refused. It does not rebuild the knowledge graph or re-run the analysis agents; their findings from the full run are carried into the change run, labelled as such, so the inbox stays complete. It is the weekly check that nothing broke — and the run to start after a template ships.
Before you start a run
Section titled “Before you start a run”The run dialog, vercatus estimate and POST /api/estimate all say the same thing before anything starts:
which allowance the run consumes and how much is left, how many requests it makes, and roughly how long
the crawl takes.
At the limit
Section titled “At the limit”A run that would exceed the site’s allowance is refused with a clear message (402 from the API) and
nothing is started. Extra runs and extra page depth are available on request; contact support or your
account contact. Nothing is added to an invoice automatically.
If you bring your own model provider
Section titled “If you bring your own model provider”On Enterprise, a workspace can route model calls through its own Anthropic key, Bedrock or Vertex. Model spend is then billed to you by that provider, and the workspace shows it — as an estimate from token counts — on Usage and on each run.