Praes pricing guide details
Pricing Overview
Praes published pricing for an observability cockpit built for OpenClaw agents. Its site is no longer reachable: praes.app returns a deployment-not-found error at its root and at /pricing, so no current pricing can be confirmed.
We are not restating the tier prices this page previously carried. They did not match the last capture we hold of the vendor's own pricing page, and with the site down there is no way to establish which, if either, was correct. Publishing a figure for a product whose site does not resolve would be the worse error.
Plan Comparison
No plan comparison can be given. The last capture of the vendor's pricing page showed a free tier at $0 and one paid amount at $15 per month, while our stored record held different figures; neither can now be verified against the source.
| What is known | Status |
|---|---|
| Product | An observability cockpit for OpenClaw agent runs — timelines, memory context, tool calls, cost analysis, guardrail monitoring |
| Vendor site | praes.app returns DEPLOYMENT_NOT_FOUND at / and /pricing, checked 9 September 2026 |
| Current pricing | Cannot be confirmed |
| Free tier | A free tier existed in the last capture; its current status is unknown |
What This Means If You Are Evaluating It
Treat Praes as unavailable for procurement until its site returns. A vendor whose marketing site, documentation and pricing page all fail to resolve cannot be assessed on price, support commitment or roadmap, and no figure we could publish would be safe to budget against.
That is a statement about availability rather than a judgement on the product. A deployment error can be transient, a domain can lapse and be renewed, and a small team can move a site without redirecting the old one. What it does mean is that nothing on the vendor's side can currently be verified.
If You Already Use Praes
Three practical questions, in order of urgency.
Does your deployment still work? A dead marketing site does not necessarily mean a dead application, but if Praes ran as a hosted service, check whether ingestion and the dashboard still respond before assuming continuity. The site was served from Vercel, and the error returned is a platform-level DEPLOYMENT_NOT_FOUND rather than an application error, which means the hosting itself is gone rather than the app failing.
Can you export your data? Agent traces, memory context and cost history are the assets worth recovering. If an API or CLI export still functions, take a copy now rather than after a support channel is needed — and prefer a raw JSON dump over a rendered report, since a replacement tool is far more likely to ingest the former.
What is your fallback? Identify a replacement for OpenClaw agent observability and confirm it can ingest what you already collect, so a migration is a decision rather than an emergency.
What to Consider Instead
For teams that need agent observability now, the alternatives listed elsewhere on this site are a better starting point than waiting for confirmation. The capabilities to match are the ones Praes was built around: a timeline of each agent run, visibility into memory context, tool-call tracking, cost attribution per run, and guardrail results. Several tools cover that ground and publish their prices, which makes them possible to budget today.