Brainontic
Simple, honest pricing.
Start with early access. Upgrade when you’re ready. No annual contracts, no “contact sales” on the personal tier, no surprises on the invoice.
Beginner
Beginner.
For getting started with your own brain.
- Your personal knowledge OS
- 10,000 chunk knowledge base
- MCP integration with Claude & Cursor
- Connectors: files, web pages, email
- Usage included — top up anytime
Personal
Personal.
For individuals building their own brain.
- Everything in Beginner
- 50,000 chunk knowledge base
- + Connectors: OneDrive, Google Drive, Dropbox (some soon)
- Usage included — top up anytime
- Personal audit trail
Pro
Pro.
For power users running agents on real work.
- Everything in Personal
- 100,000 chunk knowledge base
- + Connectors: Notion, Telegram, Jira, Zendesk, Dynamics (soon)
- More included Usage — top up anytime
SoHo
SoHo.
For a small office with a deep, sourced brain.
- Everything in Pro
- 500,000 chunk knowledge base
- + Connector: Slack (soon)
- Agent scopes & per-agent audit trail
- Most included Usage — top up anytime
Data transparency
We won’t read your data — and we make every access visible to you. Every background process is logged in your personal audit trail. The same applies to every tier.
Questions
The obvious ones first.
If we missed yours, write us. We answer everything from real email addresses, by real humans, in real sentences.
Is my data safe?
Your data stays in your brain. We don’t train models on it. We don’t hand it to a third-party LLM provider. Every background process — ingestion, entity stitching, community detection — is logged in your personal audit trail, so you can see exactly what happened, when, and why. Audit, not opacity.
What AI tools can I connect?
Anything that speaks MCP. Today that means Claude Desktop, Cursor, Claude Code, and a growing list of MCP-aware clients. Each tool gets its own named token, with its own scope. You can also bind a token to an agent_id, so an autonomous agent’s access is logged separately from yours.
What happens when I cancel?
Cancel anytime. Your data remains accessible until the end of your billing period, then it is permanently deleted from our infrastructure.
How is this different from ChatGPT Memory or Claude Projects?
Those memories live inside the model, locked to a single vendor’s context window. They can’t be exported, audited, or shared between tools. Brainontic lives outside the model. Your knowledge is yours, portable, and works with every MCP-aware AI you choose. We feed your tools — we don’t replace them.
Can I self-host?
No. Brainontic is cloud-hosted on European infrastructure. Self-hosting is not available.
Why fail-closed?
Because a knowledge OS that invents answers is worse than no knowledge OS at all. When Brainontic doesn’t have evidence, it says so. Every answer traces back to a source. That’s the feature, not a limitation.
Brainontic
“Start with early access.”
Request early access