Four packages, each priced to your scope
Privacy, Shield, Studio and the dedicated-server package protect different things — they aren't volume tiers of one feature set.
What each package covers
Privacy
Protects what employees type into AI — in the browser, over MCP and via API.
Individual pricing — based on scope.
Shield
For websites, applications and AI agents. It decides on a request before it is processed.
Individual pricing — based on scope.
Studio — Complete
The assistant over your company's documents: knowledge base, chat and analyses.
Individual pricing — based on scope.
Dedicated server / NVIDIA DGX Spark
The whole platform on a dedicated server or directly on your premises. When all of it runs on your premises, models included, your data doesn't leave your infrastructure.
Individual pricing — based on scope.
It runs where your company needs it
Deployment options are the same for every package: with us in the EU or on your premises. The details are on the home page.
Deployment optionsWhat's ready, and how it's billed
We list only what can be deployed right now, so you don't have to rely on guesswork when deciding.
- Protection in the browser for ChatGPT, Claude and Gemini
- Connection to your tools and applications
- Custom patterns for internal data
- Kill switch
- Extension for Chrome and Edge (Manifest V3)
- MCP server for Claude Desktop and Cursor
- API for your own applications
- Protection for forms, logins and chat
- Checks on what goes into the AI and what comes out
- Event overview and export
- Widget and server-side token verification
- SDK / middleware for Python, Node.js and PHP
- Proxy in front of the model (OpenAI, Anthropic)
- MCP gateway in front of agent tools
- Knowledge base with citations
- Choice of model — public provider or local
- Questions to company data without writing queries
- Access over MCP
- Read-only database queries (SELECT, WITH)
Within a package, everything is available
Scope differs — people, domains, document volume — not features.
- Privacy
- by number of people
- Shield
- by number of domains
- Studio
- by volume of documents and questions
- Dedicated server / NVIDIA DGX Spark
- by deployment scope
- — Studio — Complete includes Privacy at no extra charge
- — Outside the package you pay for cloud storage by the GB and for tokens from our AI models; if you have your own model key, you don't pay us for tokens.
The price list is part of the collaboration proposal, not this page. We do the installation together with you, and a published number would be based on an average company, not yours. So we first measure what's really happening at your company, and send you scope and a price built on your numbers only once we have the results. Requesting a proposal doesn't commit you to anything.
Let's start with measurement. You'll see what's happening in your company today.
First you see numbers from your own operation: how many bots, how much spam and what leaves for AI. What we turn on is your call after that.
Technical detailThe widget and the extension need no change to your application's code; SDK integration means adding middleware for Node.js, Python or PHP. Watching runs in monitor mode: events are recorded, nothing is stopped.