Mode 1 — On-prem puts DocMark on your own hardware, inside your own network, with no AI anywhere in the pipeline. Your analysts author the content; DocMark renders it on-brand, verifies it locally, and nothing ever leaves your premises — air-gapped capable. Here zero-egress means exactly that: no internet, no model, nothing leaves the building. For legal, valuation, M&A — anything client-confidential.

Mode 2 — Client cloud runs DocMark in your own cloud tenant (your Azure, AWS, or GCP), managed by you. Your content stays inside your cloud boundary. You bring your own LLM subscription — the LLM provider you bring, on your own account (Anthropic/Claude today, other providers on the roadmap) — and we publish recommended options and guidance. Your team runs the /v1 REST API and the MCP server inside that boundary.

Mode 3 — DocMark cloud (app.docmark.md) is our managed SaaS on Azure: DocMark-managed AI included, zero install. Here zero-egress means something different and just as real — no PII ever reaches the execution plane. Your identity stays in the control plane; the render worker never sees it.

Not sure which fits? Start here — and see the full comparison.

Ready to dig in? Compare all three modes side by side, or talk to us about a pilot.

Ready to see your brand, verified?

A pilot puts a managed DocMark box on your network — and you watch the acceptance walkthrough pass on your own machine.

Book a pilot See pricing